initial commit

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+index_dir/* filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,2 @@
+batch_pdfs/
+Pipfile

Dockerfile

@@ -0,0 +1,21 @@
+# Use an official Python runtime as a parent image
+FROM python:3.12-slim
+
+# Set the working directory in the container
+WORKDIR /app
+
+# Copy the requirements file and install dependencies
+COPY requirements.txt ./
+RUN pip install -r requirements.txt
+
+# Copy the rest of the application code
+COPY . .
+
+# Make port 8123 available to the world outside this container
+EXPOSE 8123
+
+# Define environment variable
+ENV INDEX_DIR=./index_dir
+
+# Run mcp_server.py when the container launches
+CMD ["python", "mcp_server.py", "--host", "0.0.0.0", "--port", "8123"]

app.py

@@ -0,0 +1,76 @@
+import gradio as gr
+import os
+from langchain_community.vectorstores import Chroma
+from langchain_openai import OpenAIEmbeddings
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import PyPDFLoader
+
+# Create the pdfs directory if it doesn't exist
+if not os.path.exists("pdfs"):
+    os.makedirs("pdfs")
+
+def get_pdf_files():
+    """Gets the list of PDF files from the 'pdfs' directory."""
+    return [f for f in os.listdir("pdfs") if f.endswith(".pdf")]
+
+def index_pdfs():
+    """Indexes the PDF files in the 'pdfs' directory."""
+    pdf_files = get_pdf_files()
+    if not pdf_files:
+        return "No PDF files found in the 'pdfs' directory."
+
+    success_files = []
+    failed_files = []
+    
+    for pdf_file in pdf_files:
+        try:
+            file_path = os.path.join("pdfs", pdf_file)
+            if os.path.getsize(file_path) == 0:
+                failed_files.append(f"{pdf_file} (file is empty)")
+                continue
+
+            loader = PyPDFLoader(file_path)
+            documents = loader.load()
+            text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
+            texts = text_splitter.split_documents(documents)
+            embeddings = OpenAIEmbeddings()
+            db = Chroma.from_documents(texts, embeddings, persist_directory="./db")
+            db.persist()
+            success_files.append(pdf_file)
+        except Exception as e:
+            failed_files.append(f"{pdf_file} (Error: {e})")
+
+    status = ""
+    if success_files:
+        status += f"Successfully indexed: {', '.join(success_files)}\n"
+    if failed_files:
+        status += f"Failed to index: {', '.join(failed_files)}"
+        
+    return status if status else "No files were processed."
+
+def search(query):
+    """Searches the indexed PDFs for the given query."""
+    embeddings = OpenAIEmbeddings()
+    db = Chroma(persist_directory="./db", embedding_function=embeddings)
+    docs = db.similarity_search(query)
+    results = ""
+    for doc in docs:
+        results += f"Source: {doc.metadata['source']}\n"
+        results += f"Content: {doc.page_content}\n\n"
+    return results
+
+with gr.Blocks() as demo:
+    gr.Markdown("# Simple Semantic Search App")
+    with gr.Tab("Index PDFs"):
+        pdf_files_display = gr.Textbox(label="Available PDF Files", interactive=False, value="\n".join(get_pdf_files()))
+        index_button = gr.Button("Index PDFs")
+        index_status = gr.Textbox(label="Indexing Status", interactive=False)
+        index_button.click(index_pdfs, inputs=None, outputs=index_status)
+    with gr.Tab("Search"):
+        search_query = gr.Textbox(label="Search Query")
+        search_button = gr.Button("Search")
+        search_results = gr.Textbox(label="Search Results", interactive=False)
+        search_button.click(search, inputs=search_query, outputs=search_results)
+
+if __name__ == "__main__":
+    demo.launch()

app2.py

@@ -0,0 +1,165 @@
+import gradio as gr
+import os
+import json
+from pathlib import Path
+from typing import List, Dict
+
+import numpy as np
+from sentence_transformers import SentenceTransformer, CrossEncoder
+import faiss
+
+# --- Configuration ---
+INDEX_DIR = "./index_dir"
+EMBED_MODEL = "intfloat/e5-base-v2"
+RERANKER_MODEL = "cross-encoder/ms-marco-MiniLM-L-6-v2"
+FETCH_K = 40
+TOP_K = 8
+
+# --- Global variables to hold models and data ---
+index = None
+meta = None
+embedder = None
+reranker = None
+
+# --- Utility Functions (adapted from pdf_semsearch.py) ---
+
+def e5_prefix(text: str, is_query: bool, model_name: str) -> str:
+    """Add E5-style prefixes if using an e5 model."""
+    if "e5" in model_name.lower():
+        return f"{'query' if is_query else 'passage'}: {text}"
+    return text
+
+def read_metadata(meta_path: str) -> List[Dict]:
+    """Reads metadata from a JSONL file."""
+    out = []
+    with open(meta_path, "r", encoding="utf-8") as f:
+        for line in f:
+            out.append(json.loads(line))
+    return out
+
+def pretty_snippet(s: str, max_len: int = 320) -> str:
+    """Cleans up and truncates text for display."""
+    s = " ".join(s.split())
+    return s if len(s) <= max_len else s[: max_len - 1] + "…"
+
+# --- Model and Data Loading ---
+
+def load_models_and_data():
+    """Loads the FAISS index, metadata, and models into memory."""
+    global index, meta, embedder, reranker
+
+    index_path = os.path.join(INDEX_DIR, "faiss.index")
+    meta_path = os.path.join(INDEX_DIR, "meta.jsonl")
+
+    if not os.path.exists(index_path) or not os.path.exists(meta_path):
+        raise FileNotFoundError(
+            f"Index not found in '{INDEX_DIR}'. "
+            "Please run the indexing command from pdf_semsearch.py first."
+        )
+
+    print(f"[*] Loading FAISS index: {index_path}")
+    index = faiss.read_index(index_path)
+
+    print("[*] Loading metadata…")
+    meta = read_metadata(meta_path)
+
+    print(f"[*] Loading embedding model: {EMBED_MODEL}")
+    embedder = SentenceTransformer(EMBED_MODEL)
+
+    print(f"[*] Loading reranker model: {RERANKER_MODEL}")
+    reranker = CrossEncoder(RERANKER_MODEL)
+    print("[✓] Models and data loaded.")
+
+# --- Search Function ---
+
+def search(query: str):
+    """
+    Performs semantic search on the loaded index.
+    Takes a user query, finds relevant chunks, reranks them, and returns formatted results.
+    """
+    if not query or not query.strip():
+        return "Please enter a search query."
+
+    if not all([index, meta, embedder, reranker]):
+        return "Error: Models or data not loaded. Please check the console."
+
+    # 1. Embed the query
+    query_text = e5_prefix(query, is_query=True, model_name=EMBED_MODEL)
+    qvec = embedder.encode([query_text], normalize_embeddings=True).astype("float32")
+
+    # 2. Search the FAISS index
+    D, I = index.search(qvec, FETCH_K)
+
+    # 3. Retrieve candidates
+    candidates = []
+    for j, idx in enumerate(I[0]):
+        if idx == -1:
+            continue
+        rec = dict(meta[idx])
+        rec["ann_score"] = float(D[0][j])
+        candidates.append(rec)
+
+    if not candidates:
+        return "No results found."
+
+    # 4. Rerank the candidates
+    pairs = [(query, c["text"]) for c in candidates]
+    scores = reranker.predict(pairs)
+    for c, s in zip(candidates, scores):
+        c["rerank_score"] = float(s)
+    
+    candidates.sort(key=lambda x: x["rerank_score"], reverse=True)
+
+    # 5. Format the top results for display
+    results = candidates[:TOP_K]
+    output = f"## Results for: \"{query}\"\n\n"
+    for i, r in enumerate(results, start=1):
+        base = Path(r["doc_path"]).name
+        score = r.get("rerank_score", r["ann_score"])
+        output += (
+            f"**{i}. {base} (Page: {r['page']}, Score: {score:.3f})**\n\n"
+            f"> {pretty_snippet(r['text'])}\n\n"
+            "---"
+        )
+
+    return output
+
+# --- Gradio App ---
+
+def create_gradio_app():
+    """Creates and returns the Gradio interface."""
+    with gr.Blocks(theme=gr.themes.Soft()) as iface:
+        gr.Markdown(
+            """
+            # Semantic PDF Search
+            Enter a query to search through the indexed PDF documents.
+            The index must be created first using `pdf_semsearch.py`.
+            """
+        )
+        with gr.Row():
+            query_input = gr.Textbox(
+                label="Search Query",
+                placeholder="e.g., KRAS G12C eligibility in lung cancer",
+                lines=1,
+                scale=4,
+            )
+            search_button = gr.Button("Search", variant="primary", scale=1)
+
+        results_output = gr.Markdown(label="Search Results")
+
+        search_button.click(
+            fn=search,
+            inputs=query_input,
+            outputs=results_output,
+        )
+        query_input.submit(
+            fn=search,
+            inputs=query_input,
+            outputs=results_output,
+        )
+    return iface
+
+if __name__ == "__main__":
+    load_models_and_data()
+    app = create_gradio_app()
+    app.launch()

cbioportal_study_pmids.csv

@@ -0,0 +1,733 @@
+studyId,pmid
+all_stjude_2015,25730765
+acyc_fmi_2014,24418857
+acyc_mgh_2016,26829750
+appendiceal_msk_2022,36493333
+blca_plasmacytoid_mskcc_2016,26901067
+blca_mskcc_solit_2014,25092538
+blca_nmibc_2017,28583311
+brca_mapk_hp_msk_2021,34795269
+bowel_colitis_msk_2022,36611031
+bladder_columbia_msk_2018,29625057
+bladder_msk_2023,37682528
+bm_nsclc_mskcc_2023,37591896
+cfdna_msk_2019,31768066
+ccrcc_dfci_2019,29301960
+cervix_msk_2023,37643132
+chol_jhu_2013,24185509
+chol_nccs_2013,24185513
+chol_nus_2012,22561520
+coadread_mskcc,25164765
+cllsll_icgc_2011,22158541
+coad_caseccc_2015,25583493
+chol_msk_2018,29848569
+chol_icgc_2017,28667006
+coadread_mskresistance_2022,36355783
+cscc_dfarber_2015,25589618
+ctcl_columbia_2015,26551667
+crc_eo_2020,34405229
+crc_apc_impact_2020,32730818
+crc_nigerian_2020,34819518
+crc_dd_2022,35235413
+difg_msk_2023,37910594
+escc_ucla_2014,24686850
+hcc_msk_venturaa_2018,30052636
+gct_msk_2020,32897884
+egc_msk_tp53_ccr_2022,35377946
+gbc_mskcc_2022,36228155
+gist_msk_2022,36593350
+egc_msk_2023,37699004
+hcc_jcopo_msk_2023,37769223
+histiocytosis_cobi_msk_2019,30867592
+ihch_ismms_2015,25608663
+lgsoc_mapk_msk_2022,35443055
+luad_tsp,18948947
+lymphoma_cellline_msk_2020,33067607
+lung_msk_mind_2020,36038778
+mbc_msk_2021,33863915
+mnm_washu_2016,27959731
+metastatic_solid_tumors_mich_2017,28783718
+mixed_kunga_msk_2022,35585047
+msk_ch_ped_2021,35078859
+mtnn_msk_2022,37078708
+msk_ch_2023,38147626
+nsclc_unito_2016,27346245
+nsclc_ctdx_msk_2022,36357680
+pediatric_dkfz_2017,29489754
+paired_bladder_2022,36543146
+scco_mskcc,24658004
+sarc_mskcc,20601955
+rectal_msk_2019,31591597
+rbl_cfdna_msk_2020,32633890
+rbl_mskcc_2020,33466343
+rms_msk_2023,37315267
+sarcoma_msk_2023,37350195
+skcm_vanderbilt_mskcc_2015,32913971
+soft_tissue_msk_2023,37730754
+ucec_ccr_msk_2022,35849120
+ucec_ancestry_cds_msk_2023,37651310
+ucec_msk_2024,38653864
+urcc_mskcc_2016,27713405
+utuc_mskcc_2015,26278805
+utuc_msk_2019,32332851
+utuc_pdx_msk_2019,32332851
+plmeso_msk_2024,38630790
+pancreas_msk_2024,39214094
+lms_msk_2024,38488807
+prostate_msk_2024,38949888
+panet_msk_2018,30687805
+makeanimpact_ccr_2023,36862133
+acbc_mskcc_2015,26095796
+blca_tcga_pan_can_atlas_2018,29625048
+blca_tcga_pan_can_atlas_2018,29596782
+blca_tcga_pan_can_atlas_2018,29622463
+blca_tcga_pan_can_atlas_2018,29617662
+blca_tcga_pan_can_atlas_2018,29625055
+blca_tcga_pan_can_atlas_2018,29625050
+blca_tcga_pan_can_atlas_2018,29617662
+blca_tcga_pan_can_atlas_2018,32214244
+blca_tcga_pan_can_atlas_2018,29625049
+blca_tcga_pan_can_atlas_2018,29850653
+blca_tcga_pan_can_atlas_2018,36334560
+acc_2019,31483290
+blca_msk_tcga_2020,30290956
+pcnsl_msk_2024,38995739
+msk_ctdna_vte_2024,39147831
+cellline_ccle_broad,22460905
+ccrcc_irc_2014,24487277
+ccrcc_utokyo_2013,23797736
+coadread_genentech,22895193
+cellline_nci60,22802077
+cll_iuopa_2015,26200345
+brca_metabric,27161491
+brca_metabric,30867590
+brca_metabric,22522925
+coadread_dfci_2016,27149842
+cll_broad_2015,26466571
+brca_tcga_pan_can_atlas_2018,29625048
+brca_tcga_pan_can_atlas_2018,29596782
+brca_tcga_pan_can_atlas_2018,29622463
+brca_tcga_pan_can_atlas_2018,29617662
+brca_tcga_pan_can_atlas_2018,29625055
+brca_tcga_pan_can_atlas_2018,29625050
+brca_tcga_pan_can_atlas_2018,29617662
+brca_tcga_pan_can_atlas_2018,30643250
+brca_tcga_pan_can_atlas_2018,32214244
+brca_tcga_pan_can_atlas_2018,29625049
+brca_tcga_pan_can_atlas_2018,29850653
+brca_tcga_pan_can_atlas_2018,36334560
+cesc_tcga_pan_can_atlas_2018,29625048
+cesc_tcga_pan_can_atlas_2018,29596782
+cesc_tcga_pan_can_atlas_2018,29622463
+cesc_tcga_pan_can_atlas_2018,29617662
+cesc_tcga_pan_can_atlas_2018,29625055
+cesc_tcga_pan_can_atlas_2018,29625050
+cesc_tcga_pan_can_atlas_2018,29617662
+cesc_tcga_pan_can_atlas_2018,30643250
+cesc_tcga_pan_can_atlas_2018,32214244
+cesc_tcga_pan_can_atlas_2018,29625049
+cesc_tcga_pan_can_atlas_2018,29850653
+cesc_tcga_pan_can_atlas_2018,36334560
+chol_tcga_pan_can_atlas_2018,29625048
+chol_tcga_pan_can_atlas_2018,29596782
+chol_tcga_pan_can_atlas_2018,29622463
+chol_tcga_pan_can_atlas_2018,29617662
+chol_tcga_pan_can_atlas_2018,29625055
+chol_tcga_pan_can_atlas_2018,29625050
+chol_tcga_pan_can_atlas_2018,29617662
+chol_tcga_pan_can_atlas_2018,32214244
+chol_tcga_pan_can_atlas_2018,29625049
+chol_tcga_pan_can_atlas_2018,29850653
+chol_tcga_pan_can_atlas_2018,36334560
+ccle_broad_2019,31068700
+ccle_broad_2019,31978347
+coad_cptac_2019,31031003
+coadread_cass_2020,32888432
+cll_broad_2022,35927489
+coadread_tcga_pub,22810696
+desm_broad_2015,26343386
+dlbc_broad_2012,22343534
+cscc_hgsc_bcm_2014,25303977
+coadread_tcga_pan_can_atlas_2018,29625048
+coadread_tcga_pan_can_atlas_2018,29596782
+coadread_tcga_pan_can_atlas_2018,29622463
+coadread_tcga_pan_can_atlas_2018,29617662
+coadread_tcga_pan_can_atlas_2018,29625055
+coadread_tcga_pan_can_atlas_2018,29625050
+coadread_tcga_pan_can_atlas_2018,29617662
+coadread_tcga_pan_can_atlas_2018,30643250
+coadread_tcga_pan_can_atlas_2018,32214244
+coadread_tcga_pan_can_atlas_2018,29625049
+coadread_tcga_pan_can_atlas_2018,29850653
+coadread_tcga_pan_can_atlas_2018,36334560
+dlbc_tcga_pan_can_atlas_2018,29625048
+dlbc_tcga_pan_can_atlas_2018,29596782
+dlbc_tcga_pan_can_atlas_2018,29622463
+dlbc_tcga_pan_can_atlas_2018,29617662
+dlbc_tcga_pan_can_atlas_2018,29625055
+dlbc_tcga_pan_can_atlas_2018,29625050
+dlbc_tcga_pan_can_atlas_2018,29617662
+dlbc_tcga_pan_can_atlas_2018,32214244
+dlbc_tcga_pan_can_atlas_2018,29625049
+dlbc_tcga_pan_can_atlas_2018,29850653
+dlbc_tcga_pan_can_atlas_2018,36334560
+difg_glass_2019,31748746
+cscc_ucsf_2021,34272401
+cscc_ranson_2022,35982973
+difg_glass,35649412
+difg_glass,38117484
+es_dfarber_broad_2014,25186949
+esca_tcga_pan_can_atlas_2018,29625048
+esca_tcga_pan_can_atlas_2018,29596782
+esca_tcga_pan_can_atlas_2018,29622463
+esca_tcga_pan_can_atlas_2018,29617662
+esca_tcga_pan_can_atlas_2018,29625055
+esca_tcga_pan_can_atlas_2018,29625050
+esca_tcga_pan_can_atlas_2018,29617662
+esca_tcga_pan_can_atlas_2018,32214244
+esca_tcga_pan_can_atlas_2018,29625049
+esca_tcga_pan_can_atlas_2018,29850653
+esca_tcga_pan_can_atlas_2018,36334560
+gist_msk_2023,37477937
+kirc_tcga_pub,23792563
+hcc_msk_2024,38864854
+laml_tcga_pub,23634996
+luad_mskcc_2023_met_organotropism,37084736
+mbl_sickkids_2016,26760213
+mixed_pipseq_2017,28007021
+mds_mskcc_2020,27276561
+mds_mskcc_2020,30333627
+mds_mskcc_2020,24030381
+mixed_msk_tcga_2021,34635660
+nhl_bcgsc_2011,21796119
+nhl_bcgsc_2013,23699601
+prad_broad_2013,23622249
+prad_mskcc,20579941
+prad_tcga_pub,26544944
+pcpg_tcga_pub,28162975
+pptc_2019,31693904
+prad_cdk12_mskcc_2020,32317181
+prad_pik3r1_msk_2021,35670774
+pog570_bcgsc_2020,35121966
+prad_organoids_msk_2022,35617398
+ptad_msk_2024,38758238
+prad_msk_mdanderson_2023,38488813
+stad_tcga_pub,25079317
+rectal_msk_2022,35970919
+sarcoma_msk_2022,35705558
+hnsc_tcga_pan_can_atlas_2018,29625048
+hnsc_tcga_pan_can_atlas_2018,29596782
+hnsc_tcga_pan_can_atlas_2018,29622463
+hnsc_tcga_pan_can_atlas_2018,29617662
+hnsc_tcga_pan_can_atlas_2018,29625055
+hnsc_tcga_pan_can_atlas_2018,29625050
+hnsc_tcga_pan_can_atlas_2018,29617662
+hnsc_tcga_pan_can_atlas_2018,30643250
+hnsc_tcga_pan_can_atlas_2018,32214244
+hnsc_tcga_pan_can_atlas_2018,29625049
+hnsc_tcga_pan_can_atlas_2018,29850653
+hnsc_tcga_pan_can_atlas_2018,36334560
+kich_tcga_pan_can_atlas_2018,29625048
+kich_tcga_pan_can_atlas_2018,29596782
+kich_tcga_pan_can_atlas_2018,29622463
+kich_tcga_pan_can_atlas_2018,29617662
+kich_tcga_pan_can_atlas_2018,29625055
+kich_tcga_pan_can_atlas_2018,29625050
+kich_tcga_pan_can_atlas_2018,29617662
+kich_tcga_pan_can_atlas_2018,32214244
+kich_tcga_pan_can_atlas_2018,29625049
+kich_tcga_pan_can_atlas_2018,29850653
+kich_tcga_pan_can_atlas_2018,36334560
+kirc_tcga_pan_can_atlas_2018,29625048
+kirc_tcga_pan_can_atlas_2018,29596782
+kirc_tcga_pan_can_atlas_2018,29622463
+kirc_tcga_pan_can_atlas_2018,29617662
+kirc_tcga_pan_can_atlas_2018,29625055
+kirc_tcga_pan_can_atlas_2018,29625050
+kirc_tcga_pan_can_atlas_2018,29617662
+kirc_tcga_pan_can_atlas_2018,30643250
+kirc_tcga_pan_can_atlas_2018,32214244
+kirc_tcga_pan_can_atlas_2018,29625049
+kirc_tcga_pan_can_atlas_2018,29850653
+kirc_tcga_pan_can_atlas_2018,36334560
+kirp_tcga_pan_can_atlas_2018,29625048
+kirp_tcga_pan_can_atlas_2018,29596782
+kirp_tcga_pan_can_atlas_2018,29622463
+kirp_tcga_pan_can_atlas_2018,29617662
+kirp_tcga_pan_can_atlas_2018,29625055
+kirp_tcga_pan_can_atlas_2018,29625050
+kirp_tcga_pan_can_atlas_2018,29617662
+kirp_tcga_pan_can_atlas_2018,30643250
+kirp_tcga_pan_can_atlas_2018,32214244
+kirp_tcga_pan_can_atlas_2018,29625049
+kirp_tcga_pan_can_atlas_2018,29850653
+kirp_tcga_pan_can_atlas_2018,36334560
+laml_tcga_pan_can_atlas_2018,29625048
+laml_tcga_pan_can_atlas_2018,29596782
+laml_tcga_pan_can_atlas_2018,29622463
+laml_tcga_pan_can_atlas_2018,29617662
+laml_tcga_pan_can_atlas_2018,29625055
+laml_tcga_pan_can_atlas_2018,29625050
+laml_tcga_pan_can_atlas_2018,29617662
+laml_tcga_pan_can_atlas_2018,32214244
+laml_tcga_pan_can_atlas_2018,29625049
+laml_tcga_pan_can_atlas_2018,29850653
+laml_tcga_pan_can_atlas_2018,36334560
+lihc_tcga_pan_can_atlas_2018,29625048
+lihc_tcga_pan_can_atlas_2018,29596782
+lihc_tcga_pan_can_atlas_2018,29622463
+lihc_tcga_pan_can_atlas_2018,29617662
+lihc_tcga_pan_can_atlas_2018,29625055
+lihc_tcga_pan_can_atlas_2018,29625050
+lihc_tcga_pan_can_atlas_2018,29617662
+lihc_tcga_pan_can_atlas_2018,30643250
+lihc_tcga_pan_can_atlas_2018,32214244
+lihc_tcga_pan_can_atlas_2018,29625049
+lihc_tcga_pan_can_atlas_2018,29850653
+lihc_tcga_pan_can_atlas_2018,36334560
+luad_tcga_pan_can_atlas_2018,29625048
+luad_tcga_pan_can_atlas_2018,29596782
+luad_tcga_pan_can_atlas_2018,29622463
+luad_tcga_pan_can_atlas_2018,29617662
+luad_tcga_pan_can_atlas_2018,29625055
+luad_tcga_pan_can_atlas_2018,29625050
+luad_tcga_pan_can_atlas_2018,29617662
+luad_tcga_pan_can_atlas_2018,30643250
+luad_tcga_pan_can_atlas_2018,32214244
+luad_tcga_pan_can_atlas_2018,29625049
+luad_tcga_pan_can_atlas_2018,29850653
+luad_tcga_pan_can_atlas_2018,36334560
+lusc_tcga_pan_can_atlas_2018,29625048
+lusc_tcga_pan_can_atlas_2018,29596782
+lusc_tcga_pan_can_atlas_2018,29622463
+lusc_tcga_pan_can_atlas_2018,29617662
+lusc_tcga_pan_can_atlas_2018,29625055
+lusc_tcga_pan_can_atlas_2018,29625050
+lusc_tcga_pan_can_atlas_2018,29617662
+lusc_tcga_pan_can_atlas_2018,30643250
+lusc_tcga_pan_can_atlas_2018,32214244
+lusc_tcga_pan_can_atlas_2018,29625049
+lusc_tcga_pan_can_atlas_2018,29850653
+lusc_tcga_pan_can_atlas_2018,36334560
+meso_tcga_pan_can_atlas_2018,29625048
+meso_tcga_pan_can_atlas_2018,29596782
+meso_tcga_pan_can_atlas_2018,29622463
+meso_tcga_pan_can_atlas_2018,29617662
+meso_tcga_pan_can_atlas_2018,29625055
+meso_tcga_pan_can_atlas_2018,29625050
+meso_tcga_pan_can_atlas_2018,29617662
+meso_tcga_pan_can_atlas_2018,32214244
+meso_tcga_pan_can_atlas_2018,29625049
+meso_tcga_pan_can_atlas_2018,29850653
+meso_tcga_pan_can_atlas_2018,36334560
+ov_tcga_pan_can_atlas_2018,29625048
+ov_tcga_pan_can_atlas_2018,29596782
+ov_tcga_pan_can_atlas_2018,29622463
+ov_tcga_pan_can_atlas_2018,29617662
+ov_tcga_pan_can_atlas_2018,29625055
+ov_tcga_pan_can_atlas_2018,29625050
+ov_tcga_pan_can_atlas_2018,29617662
+ov_tcga_pan_can_atlas_2018,30643250
+ov_tcga_pan_can_atlas_2018,32214244
+ov_tcga_pan_can_atlas_2018,29625049
+ov_tcga_pan_can_atlas_2018,29850653
+ov_tcga_pan_can_atlas_2018,36334560
+paad_tcga_pan_can_atlas_2018,29625048
+paad_tcga_pan_can_atlas_2018,29596782
+paad_tcga_pan_can_atlas_2018,29622463
+paad_tcga_pan_can_atlas_2018,29617662
+paad_tcga_pan_can_atlas_2018,29625055
+paad_tcga_pan_can_atlas_2018,29625050
+paad_tcga_pan_can_atlas_2018,29617662
+paad_tcga_pan_can_atlas_2018,30643250
+paad_tcga_pan_can_atlas_2018,32214244
+paad_tcga_pan_can_atlas_2018,29625049
+paad_tcga_pan_can_atlas_2018,29850653
+paad_tcga_pan_can_atlas_2018,36334560
+pcpg_tcga_pan_can_atlas_2018,29625048
+pcpg_tcga_pan_can_atlas_2018,29596782
+pcpg_tcga_pan_can_atlas_2018,29622463
+pcpg_tcga_pan_can_atlas_2018,29617662
+pcpg_tcga_pan_can_atlas_2018,29625055
+pcpg_tcga_pan_can_atlas_2018,29625050
+pcpg_tcga_pan_can_atlas_2018,29617662
+pcpg_tcga_pan_can_atlas_2018,30643250
+pcpg_tcga_pan_can_atlas_2018,32214244
+pcpg_tcga_pan_can_atlas_2018,29625049
+pcpg_tcga_pan_can_atlas_2018,29850653
+pcpg_tcga_pan_can_atlas_2018,36334560
+prad_tcga_pan_can_atlas_2018,29625048
+prad_tcga_pan_can_atlas_2018,29596782
+prad_tcga_pan_can_atlas_2018,29622463
+prad_tcga_pan_can_atlas_2018,29617662
+prad_tcga_pan_can_atlas_2018,29625055
+prad_tcga_pan_can_atlas_2018,29625050
+prad_tcga_pan_can_atlas_2018,29617662
+prad_tcga_pan_can_atlas_2018,30643250
+prad_tcga_pan_can_atlas_2018,32214244
+prad_tcga_pan_can_atlas_2018,29625049
+prad_tcga_pan_can_atlas_2018,29850653
+prad_tcga_pan_can_atlas_2018,36334560
+sarc_tcga_pan_can_atlas_2018,29625048
+sarc_tcga_pan_can_atlas_2018,29596782
+sarc_tcga_pan_can_atlas_2018,29622463
+sarc_tcga_pan_can_atlas_2018,29617662
+sarc_tcga_pan_can_atlas_2018,29625055
+sarc_tcga_pan_can_atlas_2018,29625050
+sarc_tcga_pan_can_atlas_2018,29617662
+sarc_tcga_pan_can_atlas_2018,32214244
+sarc_tcga_pan_can_atlas_2018,29625049
+sarc_tcga_pan_can_atlas_2018,29850653
+sarc_tcga_pan_can_atlas_2018,36334560
+skcm_tcga_pan_can_atlas_2018,29625048
+skcm_tcga_pan_can_atlas_2018,29596782
+skcm_tcga_pan_can_atlas_2018,29622463
+skcm_tcga_pan_can_atlas_2018,29617662
+skcm_tcga_pan_can_atlas_2018,29625055
+skcm_tcga_pan_can_atlas_2018,29625050
+skcm_tcga_pan_can_atlas_2018,29617662
+skcm_tcga_pan_can_atlas_2018,30643250
+skcm_tcga_pan_can_atlas_2018,32214244
+skcm_tcga_pan_can_atlas_2018,29625049
+skcm_tcga_pan_can_atlas_2018,29850653
+skcm_tcga_pan_can_atlas_2018,36334560
+stad_tcga_pan_can_atlas_2018,29625048
+stad_tcga_pan_can_atlas_2018,29596782
+stad_tcga_pan_can_atlas_2018,29622463
+stad_tcga_pan_can_atlas_2018,29617662
+stad_tcga_pan_can_atlas_2018,29625055
+stad_tcga_pan_can_atlas_2018,29625050
+stad_tcga_pan_can_atlas_2018,29617662
+stad_tcga_pan_can_atlas_2018,32214244
+stad_tcga_pan_can_atlas_2018,29625049
+stad_tcga_pan_can_atlas_2018,29850653
+stad_tcga_pan_can_atlas_2018,36334560
+tgct_tcga_pan_can_atlas_2018,29625048
+tgct_tcga_pan_can_atlas_2018,29596782
+tgct_tcga_pan_can_atlas_2018,29622463
+tgct_tcga_pan_can_atlas_2018,29617662
+tgct_tcga_pan_can_atlas_2018,29625055
+tgct_tcga_pan_can_atlas_2018,29625050
+tgct_tcga_pan_can_atlas_2018,29617662
+tgct_tcga_pan_can_atlas_2018,32214244
+tgct_tcga_pan_can_atlas_2018,29625049
+tgct_tcga_pan_can_atlas_2018,29850653
+tgct_tcga_pan_can_atlas_2018,36334560
+thca_tcga_pan_can_atlas_2018,29625048
+thca_tcga_pan_can_atlas_2018,29596782
+thca_tcga_pan_can_atlas_2018,29622463
+thca_tcga_pan_can_atlas_2018,29617662
+thca_tcga_pan_can_atlas_2018,29625055
+thca_tcga_pan_can_atlas_2018,29625050
+thca_tcga_pan_can_atlas_2018,29617662
+thca_tcga_pan_can_atlas_2018,30643250
+thca_tcga_pan_can_atlas_2018,32214244
+thca_tcga_pan_can_atlas_2018,29625049
+thca_tcga_pan_can_atlas_2018,29850653
+thca_tcga_pan_can_atlas_2018,36334560
+thym_tcga_pan_can_atlas_2018,29625048
+thym_tcga_pan_can_atlas_2018,29596782
+thym_tcga_pan_can_atlas_2018,29622463
+thym_tcga_pan_can_atlas_2018,29617662
+thym_tcga_pan_can_atlas_2018,29625055
+thym_tcga_pan_can_atlas_2018,29625050
+thym_tcga_pan_can_atlas_2018,29617662
+thym_tcga_pan_can_atlas_2018,32214244
+thym_tcga_pan_can_atlas_2018,29625049
+thym_tcga_pan_can_atlas_2018,29850653
+thym_tcga_pan_can_atlas_2018,36334560
+ucec_tcga_pan_can_atlas_2018,29625048
+ucec_tcga_pan_can_atlas_2018,29596782
+ucec_tcga_pan_can_atlas_2018,29622463
+ucec_tcga_pan_can_atlas_2018,29617662
+ucec_tcga_pan_can_atlas_2018,29625055
+ucec_tcga_pan_can_atlas_2018,29625050
+ucec_tcga_pan_can_atlas_2018,29617662
+ucec_tcga_pan_can_atlas_2018,30643250
+ucec_tcga_pan_can_atlas_2018,32214244
+ucec_tcga_pan_can_atlas_2018,29625049
+ucec_tcga_pan_can_atlas_2018,29850653
+ucec_tcga_pan_can_atlas_2018,36334560
+ucs_tcga_pan_can_atlas_2018,29625048
+ucs_tcga_pan_can_atlas_2018,29596782
+ucs_tcga_pan_can_atlas_2018,29622463
+ucs_tcga_pan_can_atlas_2018,29617662
+ucs_tcga_pan_can_atlas_2018,29625055
+ucs_tcga_pan_can_atlas_2018,29625050
+ucs_tcga_pan_can_atlas_2018,29617662
+ucs_tcga_pan_can_atlas_2018,32214244
+ucs_tcga_pan_can_atlas_2018,29625049
+ucs_tcga_pan_can_atlas_2018,29850653
+ucs_tcga_pan_can_atlas_2018,36334560
+uvm_tcga_pan_can_atlas_2018,29625048
+uvm_tcga_pan_can_atlas_2018,29596782
+uvm_tcga_pan_can_atlas_2018,29622463
+uvm_tcga_pan_can_atlas_2018,29617662
+uvm_tcga_pan_can_atlas_2018,29625055
+uvm_tcga_pan_can_atlas_2018,29625050
+uvm_tcga_pan_can_atlas_2018,29617662
+uvm_tcga_pan_can_atlas_2018,32214244
+uvm_tcga_pan_can_atlas_2018,29625049
+uvm_tcga_pan_can_atlas_2018,29850653
+uvm_tcga_pan_can_atlas_2018,36334560
+coad_silu_2022,37202560
+acc_tcga_pan_can_atlas_2018,29625048
+acc_tcga_pan_can_atlas_2018,29596782
+acc_tcga_pan_can_atlas_2018,29622463
+acc_tcga_pan_can_atlas_2018,29617662
+acc_tcga_pan_can_atlas_2018,29625055
+acc_tcga_pan_can_atlas_2018,29625050
+acc_tcga_pan_can_atlas_2018,29617662
+acc_tcga_pan_can_atlas_2018,32214244
+acc_tcga_pan_can_atlas_2018,29625049
+acc_tcga_pan_can_atlas_2018,29850653
+acc_tcga_pan_can_atlas_2018,36334560
+msk_chord_2024,39506116
+pancan_mappyacts_2022,35292802
+msk_met_2021,35120664
+blca_msk_2024,39499893
+brca_fuscc_2020,32719455
+thyroid_gatci_2024,38412093
+braf_msk_impact_2024,38922339
+bcc_unige_2016,26950094
+ampca_bcm_2016,26804919
+blca_dfarber_mskcc_2014,25096233
+blca_mskcc_solit_2012,23897969
+blca_bgi,24121792
+all_stjude_2013,23334668
+acyc_mskcc_2013,23685749
+acyc_jhu_2016,26862087
+acyc_mda_2015,26631609
+acyc_sanger_2013,23778141
+all_stjude_2016,27776115
+angs_project_painter_2018,32042194
+bfn_duke_nus_2015,26437033
+blca_cornell_2016,27749842
+aml_ohsu_2018,30333627
+blca_bcan_hcrn_2022,36333289
+aml_ohsu_2022,35868306
+asclc_msk_2024,39185963
+brca_bccrc_xenograft_2014,25470049
+brca_broad,22722202
+brca_bccrc,22495314
+brca_igr_2015,28027327
+blca_tcga_pub_2017,28988769
+brca_mskcc_2019,31552290
+brca_jup_msk_2020,33263939
+brain_cptac_2020,33242424
+brca_cptac_2020,33212010
+brca_dfci_2020,32404308
+brca_sanger,22722201
+brca_tcga_pub,23000897
+breast_msk_2018,30205045
+breast_alpelisib_2020,32864625
+brca_smc_2018,29713003
+breast_ink4_msk_2021,34544752
+brca_pareja_msk_2020,32220886
+crc_msk_2017,29316426
+pancan_pcawg_2020,32025007
+pdac_msk_2024,39753968
+braf_msk_archer_2024,38922339
+sarcoma_ucla_2024,39305899
+csf_msk_2024,39289779
+normal_skin_fibroblast_2024,39091884
+normal_skin_keratinocytes_2024,39091884
+normal_skin_melanocytes_2024,33029006
+normal_skin_melanocytes_2024,39091884
+normal_skin_melanocytes_2024,38895302
+normal_skin_melanocytes_2024,39975212
+chl_sccc_2023,36723991
+blca_msk_2025,40256659
+esca_broad,23525077
+escc_icgc,24670651
+es_iocurie_2014,25223734
+gbc_shanghai_2014,24997986
+egc_tmucih_2015,25583476
+egc_msk_2017,29122777
+dlbcl_duke_2017,28985567
+dlbcl_dfci_2018,29713087
+gbc_msk_2018,30427539
+egc_trap_msk_2020,32437664
+egc_mskcc_2020,33795256
+egc_trap_ccr_msk_2023,37406106
+hnsc_broad,21798893
+hnc_mskcc_2016,27442865
+hcc_inserm_fr_2015,25822088
+gct_msk_2016,27646943
+hcc_mskimpact_2018,30373752
+glioma_mskcc_2019,31263031
+glioma_msk_2018,30675060
+hccihch_pku_2019,31130341
+hgsoc_msk_2021,34819508
+hcc_meric_2021,35508466
+hcc_clca_2024,38355797
+kirc_bgi,22138691
+kich_tcga_pub,25155756
+hnsc_jhu,21798897
+hnsc_mdanderson_2013,23619168
+ihch_smmu_2014,25526346
+ihch_mskcc_2020,33963001
+ihch_msk_2021,33765338
+lgg_ucsf_2014,24336570
+lgggbm_tcga_pub,26824661
+lihc_amc_prv,24798001
+lihc_riken,22634756
+luad_mskcc_2015,25765070
+luad_broad,22980975
+liad_inserm_fr_2014,24735922
+lcll_broad_2013,23415222
+luad_msk_npjpo_2021,34290393
+luad_cptac_2020,32649874
+lusc_tcga_pub,22960745
+mbl_broad_2012,22820256
+mbl_icgc,22832583
+mbl_pcgp,22722829
+lung_msk_2017,28336552
+luad_mskcc_2020,32791233
+luad_oncosg_2020,32015526
+lung_smc_2016,27634761
+lung_pdx_msk_2021,35440124
+mbl_dkfz_2017,28726821
+lusc_cptac_2021,34358469
+lung_nci_2022,34493867
+mm_broad,24434212
+mcl_idibips_2013,24145436
+mds_tokyo_2011,21909114
+mel_tsam_liang_2017,28373299
+mel_ucla_2016,26997480
+mixed_allen_2018,30150660
+mixed_selpercatinib_2020,35304457
+mixed_cfdna_msk_2020,34059130
+mel_dfci_2019,31792460
+mel_mskimpact_2020,33509808
+mbn_sfu_2023,36201743
+mbn_msk_2024,38497151
+npc_nusingapore,24952746
+nepc_wcm_2016,26855148
+nbl_ucologne_2015,26466568
+nbl_broad_2013,23334666
+mrt_bcgsc_2016,26977886
+mpn_cimr_2013,24325359
+nsclc_mskcc_2015,25765070
+nsclc_mskcc_2018,29657128
+msk_access_2021,34145282
+mng_utoronto_2021,34433969
+mpnst_mskcc,25240281
+nbl_amc_2012,22367537
+nccrcc_genentech_2014,25401301
+ov_tcga_pub,21720365
+paac_jhu_2014,24293293
+paad_icgc,23103869
+paad_utsw_2015,25855536
+nsclc_tcga_broad_2016,27158780
+paad_qcmg_uq_2016,26909576
+pact_jhu_2011,22158988
+nsclc_tracerx_2017,28445112
+nsclc_tracerx_2017,28445469
+nsclc_pd1_msk_2018,29337640
+ntrk_msk_2019,31871300
+pan_origimed_2020,35871175
+paad_cptac_2021,34534465
+nst_nfosi_ntap,32561749
+panet_jhu_2011,21252315
+pcnsl_mayo_2015,25991819
+prad_broad,22610119
+crc_hta11_htan_2021,34910928
+panet_shanghai_2013,24326773
+plmeso_nyu_2015,25488749
+prad_cpcg_2017,28068672
+panet_arcnet_2017,28199314
+past_dkfz_heidelberg_2013,23817572
+prad_eururol_2017,28927585
+prad_fhcrc,26928463
+prad_mich,22722839
+prad_mskcc_2014,25024180
+prad_su2c_2015,26000489
+prad_mskcc_2017,28825054
+prad_p1000,29610475
+prad_su2c_2019,31061129
+prostate_dkfz_2018,30537516
+prad_msk_2019,31564440
+prad_mskcc_cheny1_organoids_2014,25201530
+prad_mcspc_mskcc_2020,32220891
+prad_msk_stopsack_2021,34667026
+prostate_pcbm_swiss_2019,35504881
+sclc_clcgp,22941188
+sclc_jhu,22941189
+skcm_broad,22817889
+rms_nih_2014,24436047
+sarc_tcga_pub,29100075
+sclc_cancercell_gardner_2017,28196596
+sclc_ucologne_2015,26168399
+sarcoma_mskcc_2022,35705560
+skcm_broad_dfarber,22622578
+skcm_yale,22842228
+stad_pfizer_uhongkong,24816253
+skcm_broad_brafresist_2012,24265153
+skcm_mskcc_2014,25409260
+skcm_tcga_pub_2015,26091043
+skcm_dfci_2015,26359337
+stad_uhongkong,22037554
+stad_utokyo,24816255
+tet_nci_2014,24974848
+thyroid_mskcc_2016,26878173
+stes_tcga_pub,28052061
+summit_2018,29420467
+stmyec_wcm_2022,36577525
+ucs_jhu_2014,25233892
+ucec_tcga_pub,23636398
+um_qimr_2016,26683228
+ucec_msk_2018,30068706
+uccc_nih_2017,28485815
+tmb_mskcc_2018,30643254
+ucec_cptac_2020,32059776
+ucec_ccr_cfdna_msk_2022,36007103
+vsc_cuk_2018,29422544
+utuc_cornell_baylor_mdacc_2019,31278255
+usarc_msk_2020,32299819
+utuc_igbmc_2021,33397444
+lgg_tcga_pan_can_atlas_2018,29625048
+lgg_tcga_pan_can_atlas_2018,29596782
+lgg_tcga_pan_can_atlas_2018,29622463
+lgg_tcga_pan_can_atlas_2018,29617662
+lgg_tcga_pan_can_atlas_2018,29625055
+lgg_tcga_pan_can_atlas_2018,29625050
+lgg_tcga_pan_can_atlas_2018,29617662
+lgg_tcga_pan_can_atlas_2018,30643250
+lgg_tcga_pan_can_atlas_2018,32214244
+lgg_tcga_pan_can_atlas_2018,29625049
+lgg_tcga_pan_can_atlas_2018,29850653
+lgg_tcga_pan_can_atlas_2018,36334560
+crc_orion_2024,39386479
+brca_aurora_2023,36585450
+schw_ctf_synodos_2025,33025139
+ovary_geomx_gray_foundation_2024,39386723
+brca_tcga_pub2015,26451490
+hnsc_tcga_pub,25631445
+luad_tcga_pub,25079552
+thca_tcga_pub,25417114
+blca_tcga_pub,24476821
+msk_ch_2020,33106634
+msk_spectrum_tme_2022,36517593
+pancan_mimsi_msk_2024,39746944
+mel_iatlas_riaz_nivolumab_2017,29033130
+stad_oncosg_2018,29670109
+gbm_tcga_pub,18772890
+gbm_tcga_pub2013,24120142
+odg_msk_2017,28472509
+gbm_tcga_pan_can_atlas_2018,29625048
+gbm_tcga_pan_can_atlas_2018,29596782
+gbm_tcga_pan_can_atlas_2018,29622463
+gbm_tcga_pan_can_atlas_2018,29617662
+gbm_tcga_pan_can_atlas_2018,29625055
+gbm_tcga_pan_can_atlas_2018,29625050
+gbm_tcga_pan_can_atlas_2018,29617662
+gbm_tcga_pan_can_atlas_2018,30643250
+gbm_tcga_pan_can_atlas_2018,32214244
+gbm_tcga_pan_can_atlas_2018,29625049
+gbm_tcga_pan_can_atlas_2018,29850653
+gbm_tcga_pan_can_atlas_2018,36334560
+gbm_mayo_pdx_sarkaria_2019,31852831
+gbm_columbia_2019,30742119
+gbm_cptac_2021,33577785
+msk_impact_2017,28481359

index_dir/faiss.index

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bf5bafdb8fd3f9bbf8bcc66f81d773d6831262a9de7f72a9eba16985cf24a7c
+size 115104813

index_dir/meta.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ef5e13e32fdc66c13c2bf468a9601f371d7a2538430af05be29c8d4a91e242b
+size 31852250

mcp_server.py

@@ -0,0 +1,200 @@
+#!/usr/bin/env python3
+"""
+Minimal MCP server using fastmcp to expose the PDF search script as a tool.
+Uses argparse to configure the server (host, port).
+Uses environment variables for the index config (INDEX_DIR, etc.).
+"""
+
+import argparse  # <-- Added this import
+import json
+import os
+from contextlib import asynccontextmanager
+from pathlib import Path
+from typing import Annotated, Dict, List, Optional
+
+# --- Vector index ---
+import faiss
+
+# --- NLP / embeddings ---
+import numpy as np
+
+# --- MCP Server ---
+from fastmcp import Context, FastMCP
+from sentence_transformers import CrossEncoder, SentenceTransformer
+
+# ---------------------------
+# Configuration (from Environment Variables)
+# ---------------------------
+INDEX_DIR = Path(os.environ.get("INDEX_DIR", "./index_dir"))
+EMBED_MODEL = os.environ.get("EMBED_MODEL", "intfloat/e5-base-v2")
+RERANKER_MODEL = os.environ.get("RERANKER_MODEL", "cross-encoder/ms-marco-MiniLM-L-6-v2")
+FETCH_K = int(os.environ.get("FETCH_K", 40))
+
+# ---------------------------
+# Global state to hold models
+# ---------------------------
+models = {}
+
+# ---------------------------
+# Copied Utilities
+# ---------------------------
+def read_metadata(meta_path: str) -> List[Dict]:
+    out = []
+    with open(meta_path, "r", encoding="utf-8") as f:
+        for line in f:
+            out.append(json.loads(line))
+    return out
+
+def e5_prefix(text: str, is_query: bool, model_name: str) -> str:
+    if "e5" in model_name.lower():
+        return f"{'query' if is_query else 'passage'}: {text}"
+    return text
+
+# ---------------------------
+# Server Startup & Shutdown
+# ---------------------------
+@asynccontextmanager
+async def lifespan(mcp: FastMCP):
+    """
+    Handles loading the models on startup.
+    """
+    print("[*] Server starting... Loading models...")
+    index_path = INDEX_DIR / "faiss.index"
+    meta_path = INDEX_DIR / "meta.jsonl"
+
+    if not index_path.exists() or not meta_path.exists():
+        print(f"[!] ERROR: Index files not found in '{INDEX_DIR}'.")
+    else:
+        print(f"[*] Loading FAISS index: {index_path}")
+        models["index"] = faiss.read_index(str(index_path))
+
+        print(f"[*] Loading metadata: {meta_path}")
+        models["meta"] = read_metadata(str(meta_path))
+
+        print(f"[*] Loading embedding model: {EMBED_MODEL}")
+        models["embedder"] = SentenceTransformer(EMBED_MODEL)
+
+        print(f"[*] Loading reranker model: {RERANKER_MODEL}")
+        models["reranker"] = CrossEncoder(RERANKER_MODEL)
+
+        print("[✓] All models and index loaded. Server is ready.")
+
+    yield
+
+    print("[*] Server shutting down... Clearing models.")
+    models.clear()
+
+# ---------------------------
+# Create the MCP Server
+# ---------------------------
+mcp = FastMCP(
+    name="PDF Corpus Search Tool",
+    lifespan=lifespan
+)
+
+# ---------------------------
+# The MCP Tool
+# ---------------------------
+Query = Annotated[str, "The semantic search query to run against the documents."]
+TopK = Annotated[int, "The final number of results to return.", "default: 5"]
+Rerank = Annotated[bool, "Whether to use a cross-encoder to rerank results. Default: true.", "default: true"]
+
+@mcp.tool()
+def search_pdf_corpus(
+    query: Query,
+    top_k: TopK = 5,
+    rerank: Rerank = True
+) -> List[Dict]:
+    """
+    Searches a private corpus of PDF documents for relevant text chunks.
+    Use this to answer questions about specific topics found in the user's files.
+    """
+    if "index" not in models:
+        return [{"error": "Index is not loaded. Check server logs."}]
+
+    # ... (search logic remains identical to before) ...
+    # 1. Get pre-loaded assets
+    index = models["index"]
+    meta = models["meta"]
+    embedder = models["embedder"]
+
+    # 2. Embed Query
+    query_text = e5_prefix(query, is_query=True, model_name=EMBED_MODEL)
+    qvec = embedder.encode([query_text], normalize_embeddings=True).astype("float32")
+
+    # 3. FAISS Search
+    D, I = index.search(qvec, FETCH_K)
+
+    # 4. Get Candidates
+    candidates = []
+    for j, idx in enumerate(I[0]):
+        if idx == -1: continue
+        rec = dict(meta[idx])
+        rec["ann_score"] = float(D[0][j])
+        candidates.append(rec)
+
+    if not candidates:
+        return []
+
+    # 5. Optional Reranking
+    if rerank:
+        pairs = [(query, c["text"]) for c in candidates]
+        scores = models["reranker"].predict(pairs)
+        for c, s in zip(candidates, scores):
+            c["rerank_score"] = float(s)
+        candidates.sort(key=lambda x: x["rerank_score"], reverse=True)
+    else:
+        candidates.sort(key=lambda x: x["ann_score"], reverse=True)
+
+    # 6. Format and return top_k results
+    final_results = []
+    for r in candidates[:top_k]:
+        final_results.append({
+            "doc_path": r["doc_path"],
+            "page": r["page"],
+            "score": r.get("rerank_score", r["ann_score"]),
+            "text": r["text"]
+        })
+
+    return final_results
+
+# ---------------------------
+# Run the Server
+# ---------------------------
+if __name__ == "__main__":
+    # --- This is the corrected block ---
+
+    parser = argparse.ArgumentParser(description="Run the PDF Search MCP Server")
+
+    # Add arguments for server configuration
+    parser.add_argument(
+        "--host",
+        type=str,
+        default="localhost",
+        help="Host to bind the server to (default: 0.0.0.0)"
+    )
+    parser.add_argument(
+        "--port",
+        type=int,
+        default=8123,
+        help="Port to run the server on (default: 8123)"
+    )
+    parser.add_argument(
+        "--transport",
+        type=str,
+        default="http",
+        choices=["http"], # fastmcp primarily uses http
+        help="Server transport protocol (default: http)"
+    )
+
+    args = parser.parse_args()
+
+    print(f"--- Starting PDF Search MCP Server on {args.transport}://{args.host}:{args.port} ---")
+    print(f"--- Using INDEX_DIR: {INDEX_DIR.resolve()} ---")
+
+    # Pass the parsed arguments to mcp.run()
+    mcp.run(
+        transport=args.transport,
+        host=args.host,
+        port=args.port
+    )

pdf_semsearch.py

@@ -0,0 +1,337 @@
+#!/usr/bin/env python3
+"""
+Tiny CLI for open-source semantic search over PDFs.
+- Index: extract → chunk → embed → FAISS
+- Search: embed query → ANN → (optional) rerank
+
+Examples:
+  # Index all PDFs in ./pdfs into ./index_dir
+  python pdf_semsearch.py index --pdf-dir ./pdfs --index-dir ./index_dir
+
+  # Search with reranking
+  python pdf_semsearch.py search --index-dir ./index_dir -q "KRAS G12C eligibility in lung cancer" --top-k 5 --rerank
+"""
+
+import os
+import sys
+import json
+import argparse
+import hashlib
+from pathlib import Path
+from typing import List, Dict, Tuple, Optional
+
+import numpy as np
+from tqdm import tqdm
+
+# --- PDF parsing / OCR ---
+import pdfplumber
+
+# OCR is optional; only imported if --ocr is used or needed
+try:
+    from pdf2image import convert_from_path  # requires poppler
+    import pytesseract  # requires tesseract runtime
+    _OCR_AVAILABLE = True
+except Exception:
+    _OCR_AVAILABLE = False
+
+# --- NLP / embeddings ---
+import spacy
+from sentence_transformers import SentenceTransformer, CrossEncoder
+
+# --- Vector index ---
+import faiss
+
+
+# ---------------------------
+# Utilities
+# ---------------------------
+def sha1_16(s: str) -> str:
+    return hashlib.sha1(s.encode("utf-8")).hexdigest()[:16]
+
+
+def ensure_dir(p: str):
+    Path(p).mkdir(parents=True, exist_ok=True)
+
+
+def load_spacy(model: str = "en_core_web_sm"):
+    try:
+        return spacy.load(model)
+    except OSError as e:
+        print(
+            f"[!] spaCy model '{model}' not found. Install it once with:\n"
+            f"    python -m spacy download {model}\n"
+        )
+        raise e
+
+
+def e5_prefix(text: str, is_query: bool, model_name: str) -> str:
+    # Add E5-style prefixes if using an e5 model
+    if "e5" in model_name.lower():
+        return f"{'query' if is_query else 'passage'}: {text}"
+    return text  # BGE & others usually don't need prefixes
+
+
+def chunk_sentences(nlp, text: str, target_chars: int = 900, overlap: int = 120) -> List[str]:
+    """Sentence-aware chunking around target_chars with soft overlap."""
+    doc = nlp(text)
+    sents = [s.text.strip() for s in doc.sents if s.text.strip()]
+    chunks, cur, cur_len = [], [], 0
+    for s in sents:
+        if cur and cur_len + len(s) > target_chars:
+            chunk = " ".join(cur)
+            chunks.append(chunk)
+            tail = chunk[-overlap:] if overlap > 0 else ""
+            cur = [tail, s] if tail else [s]
+            cur_len = len(" ".join(cur))
+        else:
+            cur.append(s)
+            cur_len += len(s)
+    if cur:
+        chunks.append(" ".join(cur))
+    # Fallback if text had no sentence boundaries
+    if not chunks and text.strip():
+        chunks = [text[:target_chars]]
+    return chunks
+
+
+def extract_pdf_text(pdf_path: str) -> List[Tuple[int, str]]:
+    """Return [(page_num, text)] using pdfplumber only (born-digital PDFs)."""
+    pages = []
+    with pdfplumber.open(pdf_path) as pdf:
+        for i, page in enumerate(pdf.pages, start=1):
+            txt = page.extract_text() or ""
+            pages.append((i, txt))
+    return pages
+
+
+def extract_pdf_text_with_ocr(pdf_path: str, dpi: int = 300, min_len: int = 20) -> List[Tuple[int, str]]:
+    """
+    Return [(page_num, text)] using pdfplumber and selective OCR if page text is too short.
+    Requires poppler & tesseract installed.
+    """
+    if not _OCR_AVAILABLE:
+        raise RuntimeError("OCR requested but pdf2image/pytesseract not available.")
+
+    out = []
+    with pdfplumber.open(pdf_path) as pdf:
+        page_count = len(pdf.pages)
+
+    for i in range(1, page_count + 1):
+        with pdfplumber.open(pdf_path) as pdf:
+            txt = (pdf.pages[i - 1].extract_text() or "").strip()
+
+        if len(txt) >= min_len:
+            out.append((i, txt))
+            continue
+
+        # OCR fallback for this page only
+        pil = convert_from_path(pdf_path, first_page=i, last_page=i, dpi=dpi)[0]
+        ocr_txt = pytesseract.image_to_string(pil, lang="eng")
+        out.append((i, ocr_txt or ""))
+
+    return out
+
+
+def build_corpus(pdf_dir: str, use_ocr: bool, nlp, chunk_chars: int, overlap: int, min_text_len_for_ocr: int) -> List[Dict]:
+    corpus = []
+    pdf_files = sorted(Path(pdf_dir).glob("**/*.pdf"))
+    for pdf_file in tqdm(pdf_files, desc="Reading PDFs"):
+        try:
+            pages = extract_pdf_text_with_ocr(str(pdf_file), min_len=min_text_len_for_ocr) if use_ocr \
+                    else extract_pdf_text(str(pdf_file))
+        except Exception as e:
+            print(f"[!] Failed to read {pdf_file}: {e}")
+            continue
+
+        for page_num, txt in pages:
+            if not txt or not txt.strip():
+                continue
+            for idx, chunk in enumerate(chunk_sentences(nlp, txt, target_chars=chunk_chars, overlap=overlap)):
+                corpus.append({
+                    "doc_path": str(pdf_file),
+                    "page": page_num,
+                    "chunk_id": idx,
+                    "text": chunk
+                })
+    return corpus
+
+
+def write_metadata(meta_path: str, corpus: List[Dict]):
+    with open(meta_path, "w", encoding="utf-8") as f:
+        for rec in corpus:
+            f.write(json.dumps(rec, ensure_ascii=False) + "\n")
+
+
+def read_metadata(meta_path: str) -> List[Dict]:
+    out = []
+    with open(meta_path, "r", encoding="utf-8") as f:
+        for line in f:
+            out.append(json.loads(line))
+    return out
+
+
+# ---------------------------
+# Indexing
+# ---------------------------
+def cmd_index(args):
+    ensure_dir(args.index_dir)
+
+    if args.ocr and not _OCR_AVAILABLE:
+        print("[!] --ocr requested but OCR deps not available. Install poppler, tesseract, pdf2image, pytesseract.")
+        sys.exit(2)
+
+    print("[*] Loading spaCy...")
+    nlp = load_spacy("en_core_web_sm")
+
+    print("[*] Building corpus from PDFs...")
+    corpus = build_corpus(
+        pdf_dir=args.pdf_dir,
+        use_ocr=args.ocr,
+        nlp=nlp,
+        chunk_chars=args.chunk_chars,
+        overlap=args.overlap,
+        min_text_len_for_ocr=args.ocr_min_text_len
+    )
+    if not corpus:
+        print("[!] No text found. Are your PDFs scanned? Try --ocr.")
+        sys.exit(1)
+
+    meta_path = os.path.join(args.index_dir, "meta.jsonl")
+    write_metadata(meta_path, corpus)
+    print(f"[*] Wrote metadata for {len(corpus)} chunks to {meta_path}")
+
+    print(f"[*] Loading embedding model: {args.embed_model}")
+    embedder = SentenceTransformer(args.embed_model)
+
+    texts = [e5_prefix(rec["text"], is_query=False, model_name=args.embed_model) for rec in corpus]
+
+    print("[*] Encoding chunks...")
+    embeddings = embedder.encode(
+        texts,
+        batch_size=args.batch_size,
+        normalize_embeddings=True,
+        show_progress_bar=True
+    ).astype("float32")
+
+    dim = embeddings.shape[1]
+    index = faiss.IndexFlatIP(dim)  # cosine via normalized vectors + inner product
+    index.add(embeddings)
+
+    index_path = os.path.join(args.index_dir, "faiss.index")
+    faiss.write_index(index, index_path)
+    print(f"[*] Wrote FAISS index to {index_path}")
+
+    print("[✓] Indexing complete.")
+
+
+# ---------------------------
+# Searching
+# ---------------------------
+def pretty_snippet(s: str, max_len: int = 320) -> str:
+    s = " ".join(s.split())
+    return s if len(s) <= max_len else s[: max_len - 1] + "…"
+
+
+def cmd_search(args):
+    index_path = os.path.join(args.index_dir, "faiss.index")
+    meta_path = os.path.join(args.index_dir, "meta.jsonl")
+    if not os.path.exists(index_path) or not os.path.exists(meta_path):
+        print("[!] Index not found. Run 'index' first.")
+        sys.exit(1)
+
+    print(f"[*] Loading FAISS index: {index_path}")
+    index = faiss.read_index(index_path)
+
+    print("[*] Loading metadata…")
+    meta = read_metadata(meta_path)
+
+    print(f"[*] Loading embedding model: {args.embed_model}")
+    embedder = SentenceTransformer(args.embed_model)
+
+    query_text = e5_prefix(args.query, is_query=True, model_name=args.embed_model)
+    qvec = embedder.encode([query_text], normalize_embeddings=True).astype("float32")
+    D, I = index.search(qvec, args.fetch_k)
+
+    candidates = []
+    for j, idx in enumerate(I[0]):
+        if idx == -1:
+            continue
+        rec = dict(meta[idx])
+        rec["ann_score"] = float(D[0][j])
+        candidates.append(rec)
+
+    if not candidates:
+        print("[!] No results.")
+        sys.exit(0)
+
+    # Optional reranking
+    if args.rerank:
+        print(f"[*] Reranking top {len(candidates)} with {args.reranker_model}…")
+        reranker = CrossEncoder(args.reranker_model)
+        pairs = [(args.query, c["text"]) for c in candidates]
+        scores = reranker.predict(pairs)
+        for c, s in zip(candidates, scores):
+            c["rerank_score"] = float(s)
+        candidates.sort(key=lambda x: x["rerank_score"], reverse=True)
+    else:
+        candidates.sort(key=lambda x: x["ann_score"], reverse=True)
+
+    results = candidates[: args.top_k]
+
+    # Print nicely
+    print("\n=== Results ===\n")
+    for i, r in enumerate(results, start=1):
+        base = Path(r["doc_path"]).name
+        score = r.get("rerank_score", r["ann_score"])
+        print(f"{i}. {base}  p.{r['page']}  score={score:.3f}")
+        print(f"   {pretty_snippet(r['text'])}\n")
+
+    if args.jsonl:
+        out = []
+        for r in results:
+            out.append({
+                "doc_path": r["doc_path"],
+                "page": r["page"],
+                "score": r.get("rerank_score", r["ann_score"]),
+                "text": r["text"]
+            })
+        print(json.dumps(out, ensure_ascii=False, indent=2))
+
+
+# ---------------------------
+# Main (argparse)
+# ---------------------------
+def main():
+    parser = argparse.ArgumentParser(description="Tiny CLI for semantic PDF search (FAISS + Sentence-Transformers)")
+    sub = parser.add_subparsers(dest="cmd", required=True)
+
+    # index
+    p_index = sub.add_parser("index", help="Index PDFs into a FAISS index")
+    p_index.add_argument("--pdf-dir", required=True, help="Folder with PDFs")
+    p_index.add_argument("--index-dir", required=True, help="Folder to write index & metadata")
+    p_index.add_argument("--embed-model", default="intfloat/e5-base-v2", help="Sentence-Transformers model name")
+    p_index.add_argument("--batch-size", type=int, default=64, help="Embedding batch size")
+    p_index.add_argument("--chunk-chars", type=int, default=900, help="Target characters per chunk")
+    p_index.add_argument("--overlap", type=int, default=120, help="Overlap characters between chunks")
+    p_index.add_argument("--ocr", action="store_true", help="Enable OCR fallback for scan-like pages")
+    p_index.add_argument("--ocr-min-text-len", type=int, default=20, help="If page text < N chars, OCR that page")
+    p_index.set_defaults(func=cmd_index)
+
+    # search
+    p_search = sub.add_parser("search", help="Search an existing index")
+    p_search.add_argument("--index-dir", required=True, help="Folder with faiss.index and meta.jsonl")
+    p_search.add_argument("-q", "--query", required=True, help="Search query")
+    p_search.add_argument("--top-k", type=int, default=8, help="How many results to show")
+    p_search.add_argument("--fetch-k", type=int, default=40, help="First-stage ANN fetch depth (before rerank)")
+    p_search.add_argument("--embed-model", default="intfloat/e5-base-v2", help="Sentence-Transformers model name")
+    p_search.add_argument("--rerank", action="store_true", help="Enable CrossEncoder reranking")
+    p_search.add_argument("--reranker-model", default="cross-encoder/ms-marco-MiniLM-L-6-v2", help="CrossEncoder name")
+    p_search.add_argument("--jsonl", action="store_true", help="Also print results as JSON to stdout")
+    p_search.set_defaults(func=cmd_search)
+
+    args = parser.parse_args()
+    args.func(args)
+
+
+if __name__ == "__main__":
+    main()

pmc_pdf_downloader.py

@@ -0,0 +1,245 @@
+#!/usr/bin/env python3
+"""
+Download NIH/PMC PDFs in batches from a CSV produced by pmid2pmcid.py.
+
+Input CSV schema (minimum):
+  pmid,pmcid,doi,status,errmsg
+Only rows with a non-empty PMCID are attempted (NIH/PMC full text).
+
+Features
+- Batch processing with per-batch delay
+- Concurrency (threaded) with polite throttling
+- Robust URL strategy (handles /pdf/ and /pdf/<PMCID>.pdf)
+- Retries with backoff for 429/5xx
+- Resume: skips already-downloaded files unless --overwrite
+- Manifest CSV of successes/failures
+
+Examples
+  python pmc_pdf_downloader.py --in pmid_to_pmcid.csv --out-dir ./pmc_pdfs \
+      --batch-size 40 --concurrency 4 --delay 1.0 --email you@org.edu
+
+  # Overwrite existing PDFs and be extra slow/polite
+  python pmc_pdf_downloader.py --in map.csv --out-dir ./pmc_pdfs --overwrite \
+      --batch-size 20 --concurrency 2 --delay 2.0
+"""
+from __future__ import annotations
+
+import argparse
+import csv
+import os
+import time
+import sys
+import math
+import re
+import pathlib
+from typing import List, Dict, Optional, Tuple
+from concurrent.futures import ThreadPoolExecutor, as_completed
+
+import requests
+
+UA = "pmc-pdf-downloader/1.0 (+https://example.org)"
+PMC_HOSTS = [
+    "https://www.ncbi.nlm.nih.gov",
+    "https://pmc.ncbi.nlm.nih.gov",
+]
+# Strategies we try in order for each host
+PMC_PATH_PATTERNS = [
+    "/pmc/articles/{pmcid}/pdf/",             # canonical "directory" that serves main PDF
+    "/pmc/articles/{pmcid}/pdf/{pmcid}.pdf",  # explicit filename
+]
+
+def read_rows(csv_path: str) -> List[Dict[str, str]]:
+    rows = []
+    with open(csv_path, newline="", encoding="utf-8") as f:
+        rd = csv.DictReader(f)
+        if not rd.fieldnames:
+            raise ValueError("CSV appears to have no header row.")
+        for r in rd:
+            rows.append({k: (v or "").strip() for k, v in r.items()})
+    return rows
+
+def valid_pmcid(pmcid: str) -> bool:
+    # Accept forms like "PMC12345" (case-insensitive)
+    return bool(re.fullmatch(r"(?i)PMC\d+", pmcid or ""))
+
+def ensure_dir(path: str):
+    pathlib.Path(path).mkdir(parents=True, exist_ok=True)
+
+def sanitize_filename(name: str) -> str:
+    return re.sub(r"[^A-Za-z0-9._\-]+", "_", name)
+
+def pick_filename(pmcid: str, pmid: str = "", doi: str = "") -> str:
+    base = pmcid.upper() if pmcid else (pmid or "UNKNOWN")
+    return sanitize_filename(f"{base}.pdf")
+
+def stream_download(url: str, dest_path: str, timeout: int, session: requests.Session) -> Tuple[bool, str]:
+    """Stream to disk; returns (ok, message)."""
+    with session.get(url, stream=True, timeout=timeout) as r:
+        if r.status_code != 200 or "application/pdf" not in r.headers.get("Content-Type", "").lower():
+            return False, f"HTTP {r.status_code} CT={r.headers.get('Content-Type')}"
+        # Respect filename from header if present
+        cd = r.headers.get("Content-Disposition", "")
+        if "filename=" in cd:
+            # simple parse; keep extension .pdf
+            fname = cd.split("filename=")[-1].strip('"; ')
+            if fname:
+                dest_dir = os.path.dirname(dest_path)
+                dest_path = os.path.join(dest_dir, sanitize_filename(fname))
+        with open(dest_path, "wb") as f:
+            for chunk in r.iter_content(chunk_size=1024 * 256):
+                if chunk:
+                    f.write(chunk)
+    return True, "ok"
+
+def try_download_pmc_pdf(pmcid: str, out_dir: str, timeout: int, session: requests.Session) -> Tuple[bool, str, str]:
+    """
+    Attempt multiple PMC URL variants across hosts. Returns (ok, msg, final_path_or_empty).
+    """
+    for host in PMC_HOSTS:
+        for pattern in PMC_PATH_PATTERNS:
+            url = f"{host}{pattern.format(pmcid=pmcid)}"
+            target = os.path.join(out_dir, pick_filename(pmcid))
+            ok, msg = stream_download(url, target, timeout=timeout, session=session)
+            if ok:
+                return True, f"{host} {msg}", target
+    return False, "no_pdf_found", ""
+
+def polite_retry(fn, *, retries=4, backoff=1.5, initial_delay=0.0, on_retry=None):
+    def wrapper(*args, **kwargs):
+        delay = initial_delay
+        attempt = 0
+        while True:
+            try:
+                return fn(*args, **kwargs)
+            except requests.RequestException as e:
+                attempt += 1
+                if attempt > retries:
+                    raise
+                if on_retry:
+                    on_retry(attempt, e)
+                time.sleep(max(0.25, delay))
+                delay *= backoff
+    return wrapper
+
+def worker(row: Dict[str, str],
+           out_dir: str,
+           timeout: int,
+           overwrite: bool,
+           email: Optional[str]) -> Dict[str, str]:
+    pmid = row.get("pmid", "")
+    pmcid = row.get("pmcid", "")
+    doi = row.get("doi", "")
+    result = {
+        "pmid": pmid,
+        "pmcid": pmcid,
+        "doi": doi,
+        "status": "",
+        "message": "",
+        "file": "",
+    }
+
+    if not valid_pmcid(pmcid):
+        result["status"] = "skip"
+        result["message"] = "no_pmcid"
+        return result
+
+    target_path = os.path.join(out_dir, pick_filename(pmcid, pmid, doi))
+    if os.path.exists(target_path) and not overwrite:
+        result["status"] = "ok_cached"
+        result["file"] = target_path
+        return result
+
+    headers = {"User-Agent": UA}
+    if email:
+        headers["From"] = email  # be nice; some servers log contact
+
+    with requests.Session() as s:
+        s.headers.update(headers)
+
+        def _on_retry(attempt, exc):
+            # print minimal retry info
+            sys.stderr.write(f"[retry] {pmcid} attempt {attempt}: {exc}\n")
+
+        safe_download = polite_retry(
+            lambda: try_download_pmc_pdf(pmcid, out_dir, timeout, s),
+            retries=4, backoff=1.8, initial_delay=0.5, on_retry=_on_retry
+        )
+
+        try:
+            ok, msg, final_path = safe_download()
+            if ok:
+                result["status"] = "ok"
+                result["message"] = msg
+                result["file"] = final_path
+            else:
+                result["status"] = "fail"
+                result["message"] = msg
+        except Exception as e:
+            result["status"] = "error"
+            result["message"] = f"{type(e).__name__}: {e}"
+
+    return result
+
+def write_manifest(path: str, rows: List[Dict[str, str]]):
+    fieldnames = ["pmid", "pmcid", "doi", "status", "message", "file"]
+    with open(path, "w", newline="", encoding="utf-8") as f:
+        w = csv.DictWriter(f, fieldnames=fieldnames)
+        w.writeheader()
+        for r in rows:
+            w.writerow(r)
+
+def chunked(seq, size):
+    for i in range(0, len(seq), size):
+        yield seq[i:i+size]
+
+def main():
+    ap = argparse.ArgumentParser(description="Download NIH/PMC PDFs in batches from a pmid→pmcid CSV.")
+    ap.add_argument("--in", dest="in_csv", required=True, help="Input CSV (from pmid2pmcid.py)")
+    ap.add_argument("--out-dir", required=True, help="Directory to write PDFs")
+    ap.add_argument("--manifest", default="pmc_download_manifest.csv", help="CSV manifest output (default: pmc_download_manifest.csv)")
+    ap.add_argument("--batch-size", type=int, default=40, help="Items per batch (default: 40)")
+    ap.add_argument("--concurrency", type=int, default=4, help="Concurrent downloads per batch (default: 4)")
+    ap.add_argument("--delay", type=float, default=1.0, help="Seconds to sleep between batches (default: 1.0)")
+    ap.add_argument("--timeout", type=int, default=60, help="Per-request timeout seconds (default: 60)")
+    ap.add_argument("--overwrite", action="store_true", help="Overwrite existing PDFs")
+    ap.add_argument("--email", help="Contact email (sent in headers)")
+    args = ap.parse_args()
+
+    ensure_dir(args.out_dir)
+    rows = read_rows(args.in_csv)
+    # keep only rows with PMCID
+    todo = [r for r in rows if valid_pmcid(r.get("pmcid", ""))]
+
+    results: List[Dict[str, str]] = []
+    total = len(todo)
+    if total == 0:
+        print("No rows with a valid PMCID found. Nothing to download.")
+        sys.exit(0)
+
+    print(f"Found {total} entries with PMCID. Starting downloads…")
+    count = 0
+    for batch_num, batch in enumerate(chunked(todo, args.batch_size), start=1):
+        print(f"Batch {batch_num}: {len(batch)} items")
+        with ThreadPoolExecutor(max_workers=args.concurrency) as ex:
+            futs = [ex.submit(worker, r, args.out_dir, args.timeout, args.overwrite, args.email) for r in batch]
+            for fut in as_completed(futs):
+                res = fut.result()
+                results.append(res)
+                count += 1
+                if res["status"].startswith("ok"):
+                    print(f"  ✓ {res['pmcid']} → {os.path.basename(res['file'])}")
+                elif res["status"] == "skip":
+                    print(f"  - {res['pmcid']} skipped ({res['message']})")
+                else:
+                    print(f"  ✗ {res['pmcid']} ({res['message']})")
+        # polite pause between batches
+        time.sleep(max(0.0, args.delay))
+
+    write_manifest(args.manifest, results)
+    ok = sum(1 for r in results if r["status"].startswith("ok"))
+    fail = sum(1 for r in results if r["status"] in ("fail", "error"))
+    skip = sum(1 for r in results if r["status"] == "skip")
+    print(f"\nDone. ok={ok}, fail={fail}, skip={skip}. Manifest: {args.manifest}")
+
+if __name__ == "__main__":
+    main()

pmcids.txt

@@ -0,0 +1,732 @@
+PMC4553269
+PMC11834963
+PMC4767593
+PMC10043565
+PMC4827439
+PMC4312739
+PMC6007852
+PMC8602441
+PMC9825391
+PMC5890941
+PMC11233068
+PMC10435547
+PMC7061455
+PMC6035749
+PMC10644000
+PMC4013720
+PMC12468851
+PMC12468851
+PMC4189196
+PMC12541443
+PMC4313860
+PMC6642361
+PMC5628134
+PMC9827113
+PMC4359951
+PMC4878831
+PMC8634406
+PMC7680360
+PMC8613248
+PMC8906458
+PMC10841595
+PMC4070589
+PMC6063411
+PMC7685753
+PMC9197876
+PMC9772093
+PMC9807588
+PMC10852615
+PMC10581608
+PMC6438729
+PMC12463604
+PMC9582036
+PMC2694412
+PMC8057264
+PMC9586871
+PMC8052452
+PMC5217532
+PMC5995337
+PMC9117241
+PMC9983778
+PMC10480533
+PMC10875331
+PMC12433863
+PMC10338177
+PMC12541558
+PMC9882421
+PMC5699446
+PMC2911503
+PMC7385919
+PMC7476838
+PMC7796332
+PMC10309566
+PMC10756077
+PMC7446400
+PMC10511463
+PMC9529954
+PMC11149479
+PMC11108776
+PMC5059781
+PMC4675454
+PMC7181640
+PMC7181640
+PMC11216861
+PMC11419252
+PMC11096044
+PMC11371520
+PMC6345401
+PMC10330105
+PMC4676955
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC6763222
+PMC6339572
+PMC11398981
+PMC11405286
+PMC3320027
+PMC4636053
+PMC12533196
+PMC3690621
+PMC3399763
+PMC12487679
+PMC4866047
+PMC6647838
+PMC3440846
+PMC4850357
+PMC4815041
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC6697103
+PMC7339254
+PMC6768830
+PMC12545938
+PMC10084830
+PMC3401966
+PMC4589486
+PMC3309757
+PMC4367811
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC6897368
+PMC8285521
+PMC9379253
+PMC9189056
+PMC10911804
+PMC12552549
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC11095631
+PMC3771322
+PMC11326964
+PMC3767041
+PMC10391526
+PMC4936195
+PMC5180407
+PMC4979995
+PMC6280667
+PMC3837510
+PMC8505423
+PMC3210554
+PMC3744992
+PMC3690918
+PMC3198787
+PMC4695400
+PMC5643159
+PMC6880934
+PMC7572747
+PMC9438279
+PMC12504664
+PMC9299269
+PMC11101347
+PMC11094415
+PMC4170219
+PMC9801308
+PMC9200814
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC10202816
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC11655358
+PMC9394403
+PMC9147702
+PMC12088707
+PMC8027015
+PMC11077417
+PMC11371517
+PMC12448954
+PMC4982376
+PMC4238969
+PMC3753703
+PMC7512009
+PMC3919793
+PMC3708595
+PMC4818686
+PMC4807116
+PMC3999050
+PMC5144107
+PMC12534164
+PMC12458290
+PMC5549141
+PMC6280667
+PMC9636269
+PMC9378589
+PMC11726019
+PMC4864027
+PMC4148686
+PMC3863681
+PMC5189935
+PMC5687509
+PMC6757060
+PMC7869928
+PMC8143193
+PMC8077737
+PMC8815415
+PMC3428862
+PMC3465532
+PMC6327853
+PMC7450824
+PMC5928087
+PMC8831444
+PMC7367727
+PMC5765991
+PMC7025898
+PMC11835752
+PMC11371517
+PMC12318355
+PMC11406943
+PMC11291049
+PMC11291049
+PMC7581540
+PMC11291049
+PMC11185634
+PMC11839034
+PMC10150291
+PMC12008543
+PMC3678719
+PMC12524419
+PMC4264969
+PMC12363069
+PMC4313862
+PMC5813492
+PMC5659841
+PMC6613387
+PMC6636637
+PMC8229851
+PMC8228505
+PMC10502449
+PMC3415217
+PMC5253129
+PMC4587544
+PMC5477828
+PMC6689131
+PMC6753053
+PMC6457907
+PMC8317046
+PMC8613272
+PMC9068765
+PMC12548999
+PMC12432380
+PMC4160352
+PMC3162986
+PMC3858325
+PMC12468851
+PMC8282702
+PMC8713028
+PMC3998672
+PMC4754110
+PMC12452113
+PMC12529571
+PMC4993154
+PMC3557932
+PMC12261305
+PMC3575604
+PMC8295366
+PMC7373300
+PMC3466113
+PMC3413789
+PMC3662966
+PMC3412905
+PMC5482929
+PMC7704768
+PMC12443039
+PMC10937974
+PMC9018685
+PMC5905700
+PMC8475722
+PMC8432745
+PMC4241387
+PMC3831489
+PMC12550706
+PMC5378171
+PMC4808437
+PMC6119118
+PMC8933489
+PMC8165771
+PMC6898788
+PMC8046739
+PMC10023728
+PMC11215372
+PMC12468275
+PMC4777652
+PMC4881306
+PMC3682833
+PMC5094835
+PMC3966280
+PMC4993154
+PMC5953836
+PMC8213710
+PMC11604310
+PMC4249650
+PMC12508281
+PMC4489427
+PMC3163504
+PMC4048021
+PMC3530898
+PMC4403382
+PMC4884143
+PMC12553238
+PMC3248495
+PMC12553238
+PMC5812436
+PMC6075848
+PMC7124988
+PMC9308789
+PMC8654574
+PMC7305302
+PMC3144496
+PMC4558226
+PMC3673022
+PMC8941949
+PMC12056210
+PMC12404184
+PMC12513462
+PMC12533323
+PMC3951336
+PMC12508145
+PMC5045679
+PMC3396711
+PMC4121784
+PMC4484602
+PMC5558263
+PMC6107367
+PMC6561293
+PMC7444093
+PMC6949382
+PMC4237931
+PMC7334067
+PMC8776579
+PMC9065149
+PMC4915822
+PMC3557461
+PMC3600117
+PMC4462130
+PMC5693358
+PMC5313262
+PMC4861069
+PMC9200818
+PMC3367798
+PMC3432702
+PMC12468435
+PMC3947264
+PMC4315319
+PMC4580370
+PMC5054517
+PMC12489181
+PMC12523349
+PMC5705185
+PMC4767360
+PMC5651175
+PMC5808581
+PMC9808553
+PMC4354107
+PMC3704730
+PMC4826231
+PMC6279519
+PMC5587124
+PMC6365097
+PMC7233456
+PMC9852004
+PMC5903820
+PMC6611775
+PMC7367750
+PMC7780630
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC11463659
+PMC9886551
+PMC7785562
+PMC11463462
+PMC4603750
+PMC4311405
+PMC4231481
+PMC4243044
+PMC3962515
+PMC7891089
+PMC9771812
+PMC11696176
+PMC5685550
+PMC5906695
+PMC2671642
+PMC3910500
+PMC5596171
+PMC5957518
+PMC6075717
+PMC6028190
+PMC5916809
+PMC6066282
+PMC6070353
+PMC5916809
+PMC12521747
+PMC7500457
+PMC5916814
+PMC5972025
+PMC12390932
+PMC7056576
+PMC6810613
+PMC8044053
+PMC5461196

pmid2pmcid.py

@@ -0,0 +1,219 @@
+#!/usr/bin/env python3
+import argparse, csv, sys, time, requests
+from typing import Iterable, List, Dict, Optional, Tuple
+try:
+    from Bio import Entrez
+except Exception:
+    Entrez = None
+
+IDCONV_URL = "https://www.ncbi.nlm.nih.gov/pmc/utils/idconv/v1.0/"
+ELINK_URL  = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi"
+UA = "pmid2pmcid-cli/1.2 (+https://example.org)"
+
+# -------------------- IO --------------------
+def read_study_pmids_from_csv(path: str) -> List[Dict[str, str]]:
+    """
+    Expect a CSV with at least: studyId, pmid
+    Falls back to any column named pmid/PMID/id if studyId is missing.
+    Returns list of dicts: {'studyId': <str or ''>, 'pmid': <str>}
+    """
+    rows = []
+    with open(path, newline="", encoding="utf-8") as f:
+        reader = csv.DictReader(f)
+        if not reader.fieldnames:
+            raise ValueError("No header row found in CSV.")
+
+        # Discover columns
+        study_col = None
+        if "studyId" in reader.fieldnames:
+            study_col = "studyId"
+
+        pmid_col = None
+        for c in ("pmid", "PMID", "id", "Id", "ID"):
+            if c in reader.fieldnames:
+                pmid_col = c
+                break
+        if not pmid_col:
+            raise ValueError(f"No pmid-like column in {reader.fieldnames}")
+
+        for row in reader:
+            pmid = (row.get(pmid_col) or "").strip()
+            if not pmid:
+                continue
+            study = (row.get(study_col) or "").strip() if study_col else ""
+            rows.append({"studyId": study, "pmid": pmid})
+    return rows
+
+def normalize_pmid(p: str) -> str:
+    p = str(p).strip()
+    if not p:
+        return ""
+    if p.lower().startswith("pmid"):
+        p = "".join(ch for ch in p if ch.isdigit())
+    return p
+
+def unique_pmids(rows: List[Dict[str, str]]) -> List[str]:
+    seen = set()
+    out = []
+    for r in rows:
+        p = normalize_pmid(r["pmid"])
+        if p and p not in seen:
+            seen.add(p)
+            out.append(p)
+    return out
+
+# -------------------- NIH resolvers --------------------
+def idconv_batch(pmids: List[str], email: Optional[str], verbose: bool) -> Dict[str, Dict]:
+    """Return mapping {pmid(str): {pmid, pmcid, doi, status, errmsg}} for a batch."""
+    params = {
+        "ids": ",".join(pmids),
+        "format": "json",
+        "tool": "pmid2pmcid-cli",
+    }
+    if email:
+        params["email"] = email
+
+    r = requests.get(IDCONV_URL, params=params, timeout=60,
+                     headers={"User-Agent": UA})
+    r.raise_for_status()
+    j = r.json()
+    if verbose:
+        print("[idconv] records:", len(j.get("records", [])))
+
+    out: Dict[str, Dict] = {}
+    for rec in j.get("records", []):
+        # Force string key so lookup matches normalized inputs
+        pmid = str(rec.get("pmid") or rec.get("requested-id") or "").strip()
+        out[pmid] = {
+            "pmid": pmid,
+            "pmcid": rec.get("pmcid") or "",
+            "doi": rec.get("doi") or "",
+            "status": rec.get("status") or "ok",
+            "errmsg": rec.get("errmsg") or "",
+        }
+
+    # Ensure every input pmid has an entry (keys are strings)
+    for p in pmids:
+        ps = str(p).strip()
+        out.setdefault(ps, {"pmid": ps, "pmcid": "", "doi": "", "status": "", "errmsg": ""})
+    return out
+
+def resolve_idconv_all(pmids: List[str], email: Optional[str], sleep=0.34, verbose=False) -> Dict[str, Dict]:
+    out: Dict[str, Dict] = {}
+    B = 200  # NIH allows up to 200 per request
+    for i in range(0, len(pmids), B):
+        batch = pmids[i:i+B]
+        m = idconv_batch(batch, email=email, verbose=verbose)
+        out.update(m)
+        time.sleep(sleep)
+    return out
+
+def elink_pubmed_to_pmc(pmid: str, email: Optional[str], api_key: Optional[str]) -> str:
+    if Entrez:
+        if not email:
+            raise ValueError("E-utilities elink requires --email when Biopython is installed.")
+        Entrez.email = email
+        if api_key:
+            Entrez.api_key = api_key
+        h = Entrez.elink(dbfrom="pubmed", db="pmc", id=pmid, retmode="xml")
+        recs = Entrez.read(h); h.close()
+        try:
+            links = recs[0]["LinkSetDb"][0]["Link"]
+            if links:
+                return "PMC" + links[0]["Id"]
+        except Exception:
+            return ""
+        return ""
+    else:
+        params = {"dbfrom": "pubmed", "db": "pmc", "id": pmid, "retmode": "json", "tool": "pmid2pmcid-cli"}
+        if email:
+            params["email"] = email
+        r = requests.get(ELINK_URL, params=params, timeout=30, headers={"User-Agent": UA})
+        r.raise_for_status()
+        j = r.json()
+        try:
+            links = j["linksets"][0]["linksetdbs"][0]["links"]
+            if links:
+                return "PMC" + str(links[0])
+        except Exception:
+            return ""
+        return ""
+
+def resolve_pmids(pmids: List[str], email: Optional[str], force_elink: bool,
+                  fallback: bool, api_key: Optional[str], verbose: bool) -> Dict[str, Dict]:
+    """Return mapping pmid(str) -> resolved fields."""
+    pmids = [normalize_pmid(p) for p in pmids if normalize_pmid(p)]
+    mapping: Dict[str, Dict] = {p: {"pmid": p, "pmcid": "", "doi": "", "status": "", "errmsg": ""} for p in pmids}
+
+    if not force_elink:
+        idc = resolve_idconv_all(pmids, email=email, verbose=verbose)
+        mapping.update(idc)
+
+    if force_elink or fallback:
+        for p in pmids:
+            need = force_elink or not mapping[p]["pmcid"]
+            if need:
+                try:
+                    pmcid = elink_pubmed_to_pmc(p, email=email, api_key=api_key)
+                except Exception as e:
+                    pmcid = ""
+                    if verbose:
+                        print(f"[elink] {p} error: {e}")
+                if pmcid:
+                    mapping[p]["pmcid"] = pmcid
+                    mapping[p]["status"] = (mapping[p]["status"] + ";elink").strip(";")
+
+    return mapping
+
+# -------------------- CLI --------------------
+def main():
+    ap = argparse.ArgumentParser(description="Convert PMID → PMCID/DOI and keep studyId in the output when provided.")
+    g_in = ap.add_mutually_exclusive_group(required=True)
+    g_in.add_argument("--pmids", help="Comma-separated PMIDs, e.g. 29625048,37261122 (no studyId)")
+    g_in.add_argument("--in-csv", help="CSV with columns 'studyId' and 'pmid' (at minimum)")
+    ap.add_argument("--out", default="pmid_to_pmcid.csv", help="Output CSV path")
+    ap.add_argument("--email", help="Your email (recommended; passed to NIH)")
+    ap.add_argument("--api-key", help="NCBI API key (optional)")
+    ap.add_argument("--fallback-elink", action="store_true", help="If ID Converter has no PMCID, try E-utilities")
+    ap.add_argument("--force-elink", action="store_true", help="Use E-utilities for ALL IDs (skip ID Converter)")
+    ap.add_argument("--verbose", action="store_true", help="Print raw API info summary")
+    args = ap.parse_args()
+
+    # Build an input list with optional studyId
+    if args.pmids:
+        input_rows = [{"studyId": "", "pmid": p.strip()} for p in args.pmids.split(",") if p.strip()]
+    else:
+        input_rows = read_study_pmids_from_csv(args.in_csv)
+
+    if not input_rows:
+        print("No input rows found.", file=sys.stderr)
+        sys.exit(1)
+
+    # Resolve unique PMIDs once
+    pmids = unique_pmids(input_rows)
+    pmid_map = resolve_pmids(pmids, email=args.email, force_elink=args.force_elink,
+                             fallback=args.fallback_elink, api_key=args.api_key, verbose=args.verbose)
+
+    # Re-expand to one row per studyId from input
+    out_rows = []
+    for r in input_rows:
+        p = normalize_pmid(r["pmid"])
+        res = pmid_map.get(p, {"pmcid": "", "doi": "", "status": "", "errmsg": ""})
+        out_rows.append({
+            "studyId": r.get("studyId", ""),
+            "pmid": p,
+            "pmcid": res.get("pmcid", ""),
+            "doi": res.get("doi", ""),
+            "status": res.get("status", ""),
+            "errmsg": res.get("errmsg", ""),
+        })
+
+    with open(args.out, "w", newline="", encoding="utf-8") as f:
+        w = csv.DictWriter(f, fieldnames=["studyId", "pmid", "pmcid", "doi", "status", "errmsg"])
+        w.writeheader()
+        w.writerows(out_rows)
+
+    print(f"[✓] Wrote {len(out_rows)} rows to {args.out}")
+
+if __name__ == "__main__":
+    main()

pmid_to_pmcid.csv

@@ -0,0 +1,733 @@
+studyId,pmid,pmcid,doi,status,errmsg
+all_stjude_2015,25730765,PMC4553269,10.1038/ng.3230,ok,
+acyc_fmi_2014,24418857,PMC11834963,,error;elink,Identifier not found in PMC
+acyc_mgh_2016,26829750,PMC4767593,10.1038/ng.3502,ok,
+appendiceal_msk_2022,36493333,PMC10043565,10.1200/JCO.22.01392,ok,
+blca_plasmacytoid_mskcc_2016,26901067,PMC4827439,10.1038/ng.3503,ok,
+blca_mskcc_solit_2014,25092538,PMC4312739,10.1016/j.eururo.2014.06.050,ok,
+blca_nmibc_2017,28583311,PMC6007852,10.1016/j.eururo.2017.05.032,ok,
+brca_mapk_hp_msk_2021,34795269,PMC8602441,10.1038/s41467-021-27093-y,ok,
+bowel_colitis_msk_2022,36611031,PMC9825391,10.1038/s41467-022-35592-9,ok,
+bladder_columbia_msk_2018,29625057,PMC5890941,10.1016/j.cell.2018.03.017,ok,
+bladder_msk_2023,37682528,PMC11233068,10.1158/1078-0432.CCR-23-1283,ok,
+bm_nsclc_mskcc_2023,37591896,PMC10435547,10.1038/s41467-023-40793-x,ok,
+cfdna_msk_2019,31768066,PMC7061455,10.1038/s41591-019-0652-7,ok,
+ccrcc_dfci_2019,29301960,PMC6035749,10.1126/science.aan5951,ok,
+cervix_msk_2023,37643132,PMC10644000,10.1158/1078-0432.CCR-23-1078,ok,
+chol_jhu_2013,24185509,PMC4013720,10.1038/ng.2813,ok,
+chol_nccs_2013,24185513,PMC12468851,,error;elink,Identifier not found in PMC
+chol_nus_2012,22561520,PMC12468851,,error;elink,Identifier not found in PMC
+coadread_mskcc,25164765,PMC4189196,10.1186/s13059-014-0454-7,ok,
+cllsll_icgc_2011,22158541,PMC12541443,,error;elink,Identifier not found in PMC
+coad_caseccc_2015,25583493,PMC4313860,10.1073/pnas.1417064112,ok,
+chol_msk_2018,29848569,PMC6642361,10.1158/1078-0432.CCR-18-0078,ok,
+chol_icgc_2017,28667006,PMC5628134,10.1158/2159-8290.CD-17-0368,ok,
+coadread_mskresistance_2022,36355783,PMC9827113,10.1158/2159-8290.CD-22-0405,ok,
+cscc_dfarber_2015,25589618,PMC4359951,10.1158/1078-0432.CCR-14-1773,ok,
+ctcl_columbia_2015,26551667,PMC4878831,10.1038/ng.3442,ok,
+crc_eo_2020,34405229,PMC8634406,10.1093/jnci/djab124,ok,
+crc_apc_impact_2020,32730818,PMC7680360,10.1053/j.gastro.2020.07.041,ok,
+crc_nigerian_2020,34819518,PMC8613248,10.1038/s41467-021-27106-w,ok,
+crc_dd_2022,35235413,PMC8906458,10.1200/PO.21.00365,ok,
+difg_msk_2023,37910594,PMC10841595,10.1158/1078-0432.CCR-23-1180,ok,
+escc_ucla_2014,24686850,PMC4070589,10.1038/ng.2935,ok,
+hcc_msk_venturaa_2018,30052636,PMC6063411,10.1371/journal.pone.0200776,ok,
+gct_msk_2020,32897884,PMC7685753,10.1172/JCI139682,ok,
+egc_msk_tp53_ccr_2022,35377946,PMC9197876,10.1158/1078-0432.CCR-21-4016,ok,
+gbc_mskcc_2022,36228155,PMC9772093,10.1158/1078-0432.CCR-22-1954,ok,
+gist_msk_2022,36593350,PMC9807588,10.1038/s41698-022-00342-z,ok,
+egc_msk_2023,37699004,PMC10852615,10.1093/jnci/djad186,ok,
+hcc_jcopo_msk_2023,37769223,PMC10581608,10.1200/PO.23.00272,ok,
+histiocytosis_cobi_msk_2019,30867592,PMC6438729,10.1038/s41586-019-1012-y,ok,
+ihch_ismms_2015,25608663,PMC12463604,,error;elink,Identifier not found in PMC
+lgsoc_mapk_msk_2022,35443055,PMC9582036,10.1158/1078-0432.CCR-21-4183,ok,
+luad_tsp,18948947,PMC2694412,10.1038/nature07423,ok,
+lymphoma_cellline_msk_2020,33067607,PMC8057264,10.1182/blood.2020008017,ok,
+lung_msk_mind_2020,36038778,PMC9586871,10.1038/s43018-022-00416-8,ok,
+mbc_msk_2021,33863915,PMC8052452,10.1038/s41523-021-00250-8,ok,
+mnm_washu_2016,27959731,PMC5217532,10.1056/NEJMoa1605949,ok,
+metastatic_solid_tumors_mich_2017,28783718,PMC5995337,10.1038/nature23306,ok,
+mixed_kunga_msk_2022,35585047,PMC9117241,10.1038/s41467-022-30233-7,ok,
+msk_ch_ped_2021,35078859,PMC9983778,10.1158/1078-0432.CCR-21-2451,ok,
+mtnn_msk_2022,37078708,PMC10480533,10.1182/bloodadvances.2023009953,ok,
+msk_ch_2023,38147626,PMC10875331,10.1182/bloodadvances.2023011262,ok,
+nsclc_unito_2016,27346245,PMC12433863,,error;elink,Identifier not found in PMC
+nsclc_ctdx_msk_2022,36357680,PMC10338177,10.1038/s41591-022-02047-z,ok,
+pediatric_dkfz_2017,29489754,PMC12541558,,error;elink,Identifier not found in PMC
+paired_bladder_2022,36543146,PMC9882421,10.1016/j.celrep.2022.111859,ok,
+scco_mskcc,24658004,PMC5699446,10.1038/ng.2922,ok,
+sarc_mskcc,20601955,PMC2911503,10.1038/ng.619,ok,
+rectal_msk_2019,31591597,PMC7385919,10.1038/s41591-019-0584-2,ok,
+rbl_cfdna_msk_2020,32633890,PMC7476838,10.1002/cam4.3144,ok,
+rbl_mskcc_2020,33466343,PMC7796332,10.3390/cancers13010149,ok,
+rms_msk_2023,37315267,PMC10309566,10.1200/PO.22.00705,ok,
+sarcoma_msk_2023,37350195,PMC10756077,10.1002/path.6137,ok,
+skcm_vanderbilt_mskcc_2015,32913971,PMC7446400,10.1200/PO.16.00054,ok,
+soft_tissue_msk_2023,37730754,PMC10511463,10.1038/s41698-023-00445-1,ok,
+ucec_ccr_msk_2022,35849120,PMC9529954,10.1158/1078-0432.CCR-22-0713,ok,
+ucec_ancestry_cds_msk_2023,37651310,PMC11149479,10.1158/2159-8290.CD-23-0546,ok,
+ucec_msk_2024,38653864,PMC11108776,10.1038/s41591-024-02942-7,ok,
+urcc_mskcc_2016,27713405,PMC5059781,10.1038/ncomms13131,ok,
+utuc_mskcc_2015,26278805,PMC4675454,10.1016/j.eururo.2015.07.039,ok,
+utuc_msk_2019,32332851,PMC7181640,10.1038/s41467-020-15885-7,ok,
+utuc_pdx_msk_2019,32332851,PMC7181640,10.1038/s41467-020-15885-7,ok,
+plmeso_msk_2024,38630790,PMC11216861,10.1158/1078-0432.CCR-24-0085,ok,
+pancreas_msk_2024,39214094,PMC11419252,10.1016/j.ccell.2024.08.002,ok,
+lms_msk_2024,38488807,PMC11096044,10.1158/1078-0432.CCR-24-0148,ok,
+prostate_msk_2024,38949888,PMC11371520,10.1158/1078-0432.CCR-23-3403,ok,
+panet_msk_2018,30687805,PMC6345401,10.1200/PO.17.00267,ok,
+makeanimpact_ccr_2023,36862133,PMC10330105,10.1158/1078-0432.CCR-22-3247,ok,
+acbc_mskcc_2015,26095796,PMC4676955,10.1002/path.4573,ok,
+blca_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+blca_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+blca_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+blca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+blca_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+blca_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+blca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+blca_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+blca_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+blca_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+blca_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+acc_2019,31483290,PMC6763222,10.1172/JCI128227,ok,
+blca_msk_tcga_2020,30290956,PMC6339572,10.1016/j.eururo.2018.09.002,ok,
+pcnsl_msk_2024,38995739,PMC11398981,10.1158/1078-0432.CCR-24-0605,ok,
+msk_ctdna_vte_2024,39147831,PMC11405286,10.1038/s41591-024-03195-0,ok,
+cellline_ccle_broad,22460905,PMC3320027,10.1038/nature11003,ok,
+ccrcc_irc_2014,24487277,PMC4636053,10.1038/ng.2891,ok,
+ccrcc_utokyo_2013,23797736,PMC12533196,,error;elink,Identifier not found in PMC
+coadread_genentech,22895193,PMC3690621,10.1038/nature11282,ok,
+cellline_nci60,22802077,PMC3399763,10.1158/0008-5472.CAN-12-1370,ok,
+cll_iuopa_2015,26200345,PMC12487679,,error;elink,Identifier not found in PMC
+brca_metabric,27161491,PMC4866047,10.1038/ncomms11479,ok,
+brca_metabric,30867590,PMC6647838,10.1038/s41586-019-1007-8,ok,
+brca_metabric,22522925,PMC3440846,10.1038/nature10983,ok,
+coadread_dfci_2016,27149842,PMC4850357,10.1016/j.celrep.2016.03.075,ok,
+cll_broad_2015,26466571,PMC4815041,10.1038/nature15395,ok,
+brca_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+brca_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+brca_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+brca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+brca_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+brca_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+brca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+brca_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+brca_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+brca_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+brca_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+brca_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+cesc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+cesc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+cesc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+cesc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+cesc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+cesc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+cesc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+cesc_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+cesc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+cesc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+cesc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+cesc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+chol_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+chol_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+chol_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+chol_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+chol_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+chol_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+chol_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+chol_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+chol_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+chol_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+chol_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+ccle_broad_2019,31068700,PMC6697103,10.1038/s41586-019-1186-3,ok,
+ccle_broad_2019,31978347,PMC7339254,10.1016/j.cell.2019.12.023,ok,
+coad_cptac_2019,31031003,PMC6768830,10.1016/j.cell.2019.03.030,ok,
+coadread_cass_2020,32888432,PMC12545938,,error;elink,Identifier not found in PMC
+cll_broad_2022,35927489,PMC10084830,10.1038/s41588-022-01140-w,ok,
+coadread_tcga_pub,22810696,PMC3401966,10.1038/nature11252,ok,
+desm_broad_2015,26343386,PMC4589486,10.1038/ng.3382,ok,
+dlbc_broad_2012,22343534,PMC3309757,10.1073/pnas.1121343109,ok,
+cscc_hgsc_bcm_2014,25303977,PMC4367811,10.1158/1078-0432.CCR-14-1768,ok,
+coadread_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+coadread_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+coadread_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+coadread_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+coadread_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+coadread_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+coadread_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+coadread_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+coadread_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+coadread_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+coadread_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+coadread_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+dlbc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+dlbc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+dlbc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+dlbc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+dlbc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+dlbc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+dlbc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+dlbc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+dlbc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+dlbc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+dlbc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+difg_glass_2019,31748746,PMC6897368,10.1038/s41586-019-1775-1,ok,
+cscc_ucsf_2021,34272401,PMC8285521,10.1038/s41525-021-00226-4,ok,
+cscc_ranson_2022,35982973,PMC9379253,10.3389/fonc.2022.919118,ok,
+difg_glass,35649412,PMC9189056,10.1016/j.cell.2022.04.038,ok,
+difg_glass,38117484,PMC10911804,10.1158/0008-5472.CAN-23-2093,ok,
+es_dfarber_broad_2014,25186949,PMC12552549,,error;elink,Identifier not found in PMC
+esca_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+esca_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+esca_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+esca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+esca_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+esca_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+esca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+esca_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+esca_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+esca_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+esca_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+gist_msk_2023,37477937,PMC11095631,10.1158/1078-0432.CCR-23-1184,ok,
+kirc_tcga_pub,23792563,PMC3771322,10.1038/nature12222,ok,
+hcc_msk_2024,38864854,PMC11326964,10.1158/1078-0432.CCR-24-0657,ok,
+laml_tcga_pub,23634996,PMC3767041,10.1056/NEJMoa1301689,ok,
+luad_mskcc_2023_met_organotropism,37084736,PMC10391526,10.1016/j.ccell.2023.03.018,ok,
+mbl_sickkids_2016,26760213,PMC4936195,10.1038/nature16478,ok,
+mixed_pipseq_2017,28007021,PMC5180407,10.1186/s13073-016-0389-6,ok,
+mds_mskcc_2020,27276561,PMC4979995,10.1056/NEJMoa1516192,ok,
+mds_mskcc_2020,30333627,PMC6280667,10.1038/s41586-018-0623-z,ok,
+mds_mskcc_2020,24030381,PMC3837510,10.1182/blood-2013-08-518886,ok,
+mixed_msk_tcga_2021,34635660,PMC8505423,10.1038/s41523-021-00339-0,ok,
+nhl_bcgsc_2011,21796119,PMC3210554,10.1038/nature10351,ok,
+nhl_bcgsc_2013,23699601,PMC3744992,10.1182/blood-2013-02-483727,ok,
+prad_broad_2013,23622249,PMC3690918,10.1016/j.cell.2013.03.021,ok,
+prad_mskcc,20579941,PMC3198787,10.1016/j.ccr.2010.05.026,ok,
+prad_tcga_pub,26544944,PMC4695400,10.1016/j.cell.2015.10.025,ok,
+pcpg_tcga_pub,28162975,PMC5643159,10.1016/j.ccell.2017.01.001,ok,
+pptc_2019,31693904,PMC6880934,10.1016/j.celrep.2019.09.071,ok,
+prad_cdk12_mskcc_2020,32317181,PMC7572747,10.1016/j.eururo.2020.03.024,ok,
+prad_pik3r1_msk_2021,35670774,PMC9438279,10.1158/1078-0432.CCR-21-4272,ok,
+pog570_bcgsc_2020,35121966,PMC12504664,,error;elink,Identifier not found in PMC
+prad_organoids_msk_2022,35617398,PMC9299269,10.1126/science.abe1505,ok,
+ptad_msk_2024,38758238,PMC11101347,10.1007/s00401-024-02736-8,ok,
+prad_msk_mdanderson_2023,38488813,PMC11094415,10.1158/1078-0432.CCR-23-2438,ok,
+stad_tcga_pub,25079317,PMC4170219,10.1038/nature13480,ok,
+rectal_msk_2022,35970919,PMC9801308,10.1038/s41591-022-01930-z,ok,
+sarcoma_msk_2022,35705558,PMC9200814,10.1038/s41467-022-30496-0,ok,
+hnsc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+hnsc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+hnsc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+hnsc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+hnsc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+hnsc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+hnsc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+hnsc_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+hnsc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+hnsc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+hnsc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+hnsc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+kich_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+kich_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+kich_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+kich_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+kich_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+kich_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+kich_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+kich_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+kich_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+kich_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+kich_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+kirc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+kirc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+kirc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+kirc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+kirc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+kirc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+kirc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+kirc_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+kirc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+kirc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+kirc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+kirc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+kirp_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+kirp_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+kirp_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+kirp_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+kirp_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+kirp_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+kirp_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+kirp_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+kirp_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+kirp_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+kirp_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+kirp_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+laml_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+laml_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+laml_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+laml_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+laml_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+laml_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+laml_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+laml_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+laml_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+laml_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+laml_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+lihc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+lihc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+lihc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+lihc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+lihc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+lihc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+lihc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+lihc_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+lihc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+lihc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+lihc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+lihc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+luad_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+luad_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+luad_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+luad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+luad_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+luad_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+luad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+luad_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+luad_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+luad_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+luad_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+luad_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+lusc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+lusc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+lusc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+lusc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+lusc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+lusc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+lusc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+lusc_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+lusc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+lusc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+lusc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+lusc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+meso_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+meso_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+meso_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+meso_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+meso_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+meso_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+meso_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+meso_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+meso_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+meso_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+meso_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+ov_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+ov_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+ov_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+ov_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+ov_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+ov_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+ov_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+ov_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+ov_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+ov_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+ov_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+ov_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+paad_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+paad_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+paad_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+paad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+paad_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+paad_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+paad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+paad_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+paad_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+paad_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+paad_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+paad_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+pcpg_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+pcpg_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+pcpg_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+pcpg_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+pcpg_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+pcpg_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+pcpg_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+pcpg_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+pcpg_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+pcpg_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+pcpg_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+pcpg_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+prad_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+prad_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+prad_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+prad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+prad_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+prad_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+prad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+prad_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+prad_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+prad_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+prad_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+prad_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+sarc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+sarc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+sarc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+sarc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+sarc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+sarc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+sarc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+sarc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+sarc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+sarc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+sarc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+skcm_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+skcm_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+skcm_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+skcm_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+skcm_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+skcm_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+skcm_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+skcm_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+skcm_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+skcm_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+skcm_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+skcm_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+stad_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+stad_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+stad_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+stad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+stad_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+stad_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+stad_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+stad_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+stad_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+stad_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+stad_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+tgct_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+tgct_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+tgct_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+tgct_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+tgct_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+tgct_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+tgct_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+tgct_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+tgct_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+tgct_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+tgct_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+thca_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+thca_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+thca_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+thca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+thca_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+thca_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+thca_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+thca_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+thca_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+thca_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+thca_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+thca_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+thym_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+thym_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+thym_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+thym_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+thym_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+thym_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+thym_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+thym_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+thym_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+thym_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+thym_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+ucec_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+ucec_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+ucec_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+ucec_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+ucec_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+ucec_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+ucec_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+ucec_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+ucec_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+ucec_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+ucec_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+ucec_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+ucs_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+ucs_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+ucs_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+ucs_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+ucs_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+ucs_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+ucs_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+ucs_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+ucs_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+ucs_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+ucs_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+uvm_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+uvm_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+uvm_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+uvm_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+uvm_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+uvm_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+uvm_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+uvm_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+uvm_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+uvm_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+uvm_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+coad_silu_2022,37202560,PMC10202816,10.1038/s41591-023-02324-5,ok,
+acc_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+acc_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+acc_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+acc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+acc_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+acc_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+acc_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+acc_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+acc_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+acc_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+acc_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+msk_chord_2024,39506116,PMC11655358,10.1038/s41586-024-08167-5,ok,
+pancan_mappyacts_2022,35292802,PMC9394403,10.1158/2159-8290.CD-21-1136,ok,
+msk_met_2021,35120664,PMC9147702,10.1016/j.cell.2022.01.003,ok,
+blca_msk_2024,39499893,PMC12088707,10.1200/PO.24.00287,ok,
+brca_fuscc_2020,32719455,PMC8027015,10.1038/s41422-020-0375-9,ok,
+thyroid_gatci_2024,38412093,PMC11077417,10.1016/j.celrep.2024.113826,ok,
+braf_msk_impact_2024,38922339,PMC11371517,10.1158/1078-0432.CCR-23-3981,ok,
+bcc_unige_2016,26950094,PMC12448954,,error;elink,Identifier not found in PMC
+ampca_bcm_2016,26804919,PMC4982376,10.1016/j.celrep.2015.12.005,ok,
+blca_dfarber_mskcc_2014,25096233,PMC4238969,10.1158/2159-8290.CD-14-0623,ok,
+blca_mskcc_solit_2012,23897969,PMC3753703,10.1200/JCO.2012.46.5740,ok,
+blca_bgi,24121792,PMC7512009,10.1038/ng.2798,ok,
+all_stjude_2013,23334668,PMC3919793,10.1038/ng.2532,ok,
+acyc_mskcc_2013,23685749,PMC3708595,10.1038/ng.2643,ok,
+acyc_jhu_2016,26862087,PMC4818686,10.1158/1940-6207.CAPR-15-0316,ok,
+acyc_mda_2015,26631609,PMC4807116,10.1158/1078-0432.CCR-15-2867-T,ok,
+acyc_sanger_2013,23778141,PMC3999050,10.1172/JCI67201,ok,
+all_stjude_2016,27776115,PMC5144107,10.1038/ng.3691,ok,
+angs_project_painter_2018,32042194,PMC12534164,,error;elink,Identifier not found in PMC
+bfn_duke_nus_2015,26437033,PMC12458290,,error;elink,Identifier not found in PMC
+blca_cornell_2016,27749842,PMC5549141,10.1038/ng.3692,ok,
+aml_ohsu_2018,30333627,PMC6280667,10.1038/s41586-018-0623-z,ok,
+blca_bcan_hcrn_2022,36333289,PMC9636269,10.1038/s41467-022-33980-9,ok,
+aml_ohsu_2022,35868306,PMC9378589,10.1016/j.ccell.2022.07.002,ok,
+asclc_msk_2024,39185963,PMC11726019,10.1158/2159-8290.CD-24-0286,ok,
+brca_bccrc_xenograft_2014,25470049,PMC4864027,10.1038/nature13952,ok,
+brca_broad,22722202,PMC4148686,10.1038/nature11154,ok,
+brca_bccrc,22495314,PMC3863681,10.1038/nature10933,ok,
+brca_igr_2015,28027327,PMC5189935,10.1371/journal.pmed.1002201,ok,
+blca_tcga_pub_2017,28988769,PMC5687509,10.1016/j.cell.2017.09.007,ok,
+brca_mskcc_2019,31552290,PMC6757060,10.1038/s41523-019-0126-6,ok,
+brca_jup_msk_2020,33263939,PMC7869928,10.1002/cjp2.190,ok,
+brain_cptac_2020,33242424,PMC8143193,10.1016/j.cell.2020.10.044,ok,
+brca_cptac_2020,33212010,PMC8077737,10.1016/j.cell.2020.10.036,ok,
+brca_dfci_2020,32404308,PMC8815415,10.1158/2159-8290.CD-19-1390,ok,
+brca_sanger,22722201,PMC3428862,10.1038/nature11017,ok,
+brca_tcga_pub,23000897,PMC3465532,10.1038/nature11412,ok,
+breast_msk_2018,30205045,PMC6327853,10.1016/j.ccell.2018.08.008,ok,
+breast_alpelisib_2020,32864625,PMC7450824,10.1038/s43018-020-0047-1,ok,
+brca_smc_2018,29713003,PMC5928087,10.1038/s41467-018-04129-4,ok,
+breast_ink4_msk_2021,34544752,PMC8831444,10.1158/2159-8290.CD-20-1726,ok,
+brca_pareja_msk_2020,32220886,PMC7367727,10.1158/1078-0432.CCR-19-2563,ok,
+crc_msk_2017,29316426,PMC5765991,10.1016/j.ccell.2017.12.004,ok,
+pancan_pcawg_2020,32025007,PMC7025898,10.1038/s41586-020-1969-6,ok,
+pdac_msk_2024,39753968,PMC11835752,10.1038/s41591-024-03362-3,ok,
+braf_msk_archer_2024,38922339,PMC11371517,10.1158/1078-0432.CCR-23-3981,ok,
+sarcoma_ucla_2024,39305899,PMC12318355,10.1016/j.stem.2024.08.010,ok,
+csf_msk_2024,39289779,PMC11406943,10.1186/s40478-024-01846-4,ok,
+normal_skin_fibroblast_2024,39091884,PMC11291049,10.1101/2024.07.23.604673,ok,
+normal_skin_keratinocytes_2024,39091884,PMC11291049,10.1101/2024.07.23.604673,ok,
+normal_skin_melanocytes_2024,33029006,PMC7581540,10.1038/s41586-020-2785-8,ok,
+normal_skin_melanocytes_2024,39091884,PMC11291049,10.1101/2024.07.23.604673,ok,
+normal_skin_melanocytes_2024,38895302,PMC11185634,10.1101/2024.06.04.597225,ok,
+normal_skin_melanocytes_2024,39975212,PMC11839034,10.1101/2025.02.07.637114,ok,
+chl_sccc_2023,36723991,PMC10150291,10.1158/2643-3230.BCD-22-0128,ok,
+blca_msk_2025,40256659,PMC12008543,10.1016/j.euros.2025.03.009,ok,
+esca_broad,23525077,PMC3678719,10.1038/ng.2591,ok,
+escc_icgc,24670651,PMC12524419,,error;elink,Identifier not found in PMC
+es_iocurie_2014,25223734,PMC4264969,10.1158/2159-8290.CD-14-0622,ok,
+gbc_shanghai_2014,24997986,PMC12363069,,error;elink,Identifier not found in PMC
+egc_tmucih_2015,25583476,PMC4313862,10.1073/pnas.1422640112,ok,
+egc_msk_2017,29122777,PMC5813492,10.1158/2159-8290.CD-17-0787,ok,
+dlbcl_duke_2017,28985567,PMC5659841,10.1016/j.cell.2017.09.027,ok,
+dlbcl_dfci_2018,29713087,PMC6613387,10.1038/s41591-018-0016-8,ok,
+gbc_msk_2018,30427539,PMC6636637,10.1002/cncr.31850,ok,
+egc_trap_msk_2020,32437664,PMC8229851,10.1016/S1470-2045(20)30169-8,ok,
+egc_mskcc_2020,33795256,PMC8228505,10.1158/1078-0432.CCR-20-4707,ok,
+egc_trap_ccr_msk_2023,37406106,PMC10502449,10.1158/1078-0432.CCR-22-3769,ok,
+hnsc_broad,21798893,PMC3415217,10.1126/science.1208130,ok,
+hnc_mskcc_2016,27442865,PMC5253129,10.1001/jamaoncol.2016.1790,ok,
+hcc_inserm_fr_2015,25822088,PMC4587544,10.1038/ng.3252,ok,
+gct_msk_2016,27646943,PMC5477828,10.1200/JCO.2016.68.7798,ok,
+hcc_mskimpact_2018,30373752,PMC6689131,10.1158/1078-0432.CCR-18-2293,ok,
+glioma_mskcc_2019,31263031,PMC6753053,10.1158/1078-0432.CCR-19-0032,ok,
+glioma_msk_2018,30675060,PMC6457907,10.1038/s41586-019-0882-3,ok,
+hccihch_pku_2019,31130341,PMC8317046,10.1016/j.ccell.2019.04.007,ok,
+hgsoc_msk_2021,34819508,PMC8613272,10.1038/s41525-021-00259-9,ok,
+hcc_meric_2021,35508466,PMC9068765,10.1038/s41467-022-29960-8,ok,
+hcc_clca_2024,38355797,PMC12548999,,error;elink,Identifier not found in PMC
+kirc_bgi,22138691,PMC12432380,,error;elink,Identifier not found in PMC
+kich_tcga_pub,25155756,PMC4160352,10.1016/j.ccr.2014.07.014,ok,
+hnsc_jhu,21798897,PMC3162986,10.1126/science.1206923,ok,
+hnsc_mdanderson_2013,23619168,PMC3858325,10.1158/2159-8290.CD-12-0537,ok,
+ihch_smmu_2014,25526346,PMC12468851,,error;elink,Identifier not found in PMC
+ihch_mskcc_2020,33963001,PMC8282702,10.1158/1078-0432.CCR-21-0412,ok,
+ihch_msk_2021,33765338,PMC8713028,10.1002/hep.31829,ok,
+lgg_ucsf_2014,24336570,PMC3998672,10.1126/science.1239947,ok,
+lgggbm_tcga_pub,26824661,PMC4754110,10.1016/j.cell.2015.12.028,ok,
+lihc_amc_prv,24798001,PMC12452113,,error;elink,Identifier not found in PMC
+lihc_riken,22634756,PMC12529571,,error;elink,Identifier not found in PMC
+luad_mskcc_2015,25765070,PMC4993154,10.1126/science.aaa1348,ok,
+luad_broad,22980975,PMC3557932,10.1016/j.cell.2012.08.029,ok,
+liad_inserm_fr_2014,24735922,PMC12261305,,error;elink,Identifier not found in PMC
+lcll_broad_2013,23415222,PMC3575604,10.1016/j.cell.2013.01.019,ok,
+luad_msk_npjpo_2021,34290393,PMC8295366,10.1038/s41698-021-00210-2,ok,
+luad_cptac_2020,32649874,PMC7373300,10.1016/j.cell.2020.06.013,ok,
+lusc_tcga_pub,22960745,PMC3466113,10.1038/nature11404,ok,
+mbl_broad_2012,22820256,PMC3413789,10.1038/nature11329,ok,
+mbl_icgc,22832583,PMC3662966,10.1038/nature11284,ok,
+mbl_pcgp,22722829,PMC3412905,10.1038/nature11213,ok,
+lung_msk_2017,28336552,PMC5482929,10.1158/2159-8290.CD-16-1337,ok,
+luad_mskcc_2020,32791233,PMC7704768,10.1016/j.jtho.2020.08.005,ok,
+luad_oncosg_2020,32015526,PMC12443039,,error;elink,Identifier not found in PMC
+lung_smc_2016,27634761,PMC10937974,,error;elink,Identifier not found in PMC
+lung_pdx_msk_2021,35440124,PMC9018685,10.1038/s41467-022-29794-4,ok,
+mbl_dkfz_2017,28726821,PMC5905700,10.1038/nature22973,ok,
+lusc_cptac_2021,34358469,PMC8475722,10.1016/j.cell.2021.07.016,ok,
+lung_nci_2022,34493867,PMC8432745,10.1038/s41588-021-00920-0,ok,
+mm_broad,24434212,PMC4241387,10.1016/j.ccr.2013.12.015,ok,
+mcl_idibips_2013,24145436,PMC3831489,10.1073/pnas.1314608110,ok,
+mds_tokyo_2011,21909114,PMC12550706,,error;elink,Identifier not found in PMC
+mel_tsam_liang_2017,28373299,PMC5378171,10.1101/gr.213348.116,ok,
+mel_ucla_2016,26997480,PMC4808437,10.1016/j.cell.2016.02.065,ok,
+mixed_allen_2018,30150660,PMC6119118,10.1038/s41588-018-0200-2,ok,
+mixed_selpercatinib_2020,35304457,PMC8933489,10.1038/s41467-022-28848-x,ok,
+mixed_cfdna_msk_2020,34059130,PMC8165771,10.1186/s13073-021-00898-8,ok,
+mel_dfci_2019,31792460,PMC6898788,10.1038/s41591-019-0654-5,ok,
+mel_mskimpact_2020,33509808,PMC8046739,10.1158/1078-0432.CCR-20-4189,ok,
+mbn_sfu_2023,36201743,PMC10023728,10.1182/blood.2022016534,ok,
+mbn_msk_2024,38497151,PMC11215372,10.3324/haematol.2023.284565,ok,
+npc_nusingapore,24952746,PMC12468275,,error;elink,Identifier not found in PMC
+nepc_wcm_2016,26855148,PMC4777652,10.1038/nm.4045,ok,
+nbl_ucologne_2015,26466568,PMC4881306,10.1038/nature14980,ok,
+nbl_broad_2013,23334666,PMC3682833,10.1038/ng.2529,ok,
+mrt_bcgsc_2016,26977886,PMC5094835,10.1016/j.ccell.2016.02.009,ok,
+mpn_cimr_2013,24325359,PMC3966280,10.1056/NEJMoa1312542,ok,
+nsclc_mskcc_2015,25765070,PMC4993154,10.1126/science.aaa1348,ok,
+nsclc_mskcc_2018,29657128,PMC5953836,10.1016/j.ccell.2018.03.018,ok,
+msk_access_2021,34145282,PMC8213710,10.1038/s41467-021-24109-5,ok,
+mng_utoronto_2021,34433969,PMC11604310,10.1038/s41586-021-03850-3,ok,
+mpnst_mskcc,25240281,PMC4249650,10.1038/ng.3095,ok,
+nbl_amc_2012,22367537,PMC12508281,,error;elink,Identifier not found in PMC
+nccrcc_genentech_2014,25401301,PMC4489427,10.1038/ng.3146,ok,
+ov_tcga_pub,21720365,PMC3163504,10.1038/nature10166,ok,
+paac_jhu_2014,24293293,PMC4048021,10.1002/path.4310,ok,
+paad_icgc,23103869,PMC3530898,10.1038/nature11547,ok,
+paad_utsw_2015,25855536,PMC4403382,10.1038/ncomms7744,ok,
+nsclc_tcga_broad_2016,27158780,PMC4884143,10.1038/ng.3564,ok,
+paad_qcmg_uq_2016,26909576,PMC12553238,,error;elink,Identifier not found in PMC
+pact_jhu_2011,22158988,PMC3248495,10.1073/pnas.1118046108,ok,
+nsclc_tracerx_2017,28445112,PMC12553238,,error;elink,Identifier not found in PMC
+nsclc_tracerx_2017,28445469,PMC5812436,10.1038/nature22364,ok,
+nsclc_pd1_msk_2018,29337640,PMC6075848,10.1200/JCO.2017.75.3384,ok,
+ntrk_msk_2019,31871300,PMC7124988,10.1158/1078-0432.CCR-19-3165,ok,
+pan_origimed_2020,35871175,PMC9308789,10.1038/s41467-022-31780-9,ok,
+paad_cptac_2021,34534465,PMC8654574,10.1016/j.cell.2021.08.023,ok,
+nst_nfosi_ntap,32561749,PMC7305302,10.1038/s41597-020-0508-5,ok,
+panet_jhu_2011,21252315,PMC3144496,10.1126/science.1200609,ok,
+pcnsl_mayo_2015,25991819,PMC4558226,10.1158/1078-0432.CCR-14-2116,ok,
+prad_broad,22610119,PMC3673022,10.1038/ng.2279,ok,
+crc_hta11_htan_2021,34910928,PMC8941949,10.1016/j.cell.2021.11.031,ok,
+panet_shanghai_2013,24326773,PMC12056210,,error;elink,Identifier not found in PMC
+plmeso_nyu_2015,25488749,PMC12404184,,error;elink,Identifier not found in PMC
+prad_cpcg_2017,28068672,PMC12513462,,error;elink,Identifier not found in PMC
+panet_arcnet_2017,28199314,PMC12533323,,error;elink,Identifier not found in PMC
+past_dkfz_heidelberg_2013,23817572,PMC3951336,10.1038/ng.2682,ok,
+prad_eururol_2017,28927585,PMC12508145,,error;elink,Identifier not found in PMC
+prad_fhcrc,26928463,PMC5045679,10.1038/nm.4053,ok,
+prad_mich,22722839,PMC3396711,10.1038/nature11125,ok,
+prad_mskcc_2014,25024180,PMC4121784,10.1073/pnas.1411446111,ok,
+prad_su2c_2015,26000489,PMC4484602,10.1016/j.cell.2015.05.001,ok,
+prad_mskcc_2017,28825054,PMC5558263,10.1200/PO.17.00029,ok,
+prad_p1000,29610475,PMC6107367,10.1038/s41588-018-0078-z,ok,
+prad_su2c_2019,31061129,PMC6561293,10.1073/pnas.1902651116,ok,
+prostate_dkfz_2018,30537516,PMC7444093,10.1016/j.ccell.2018.10.016,ok,
+prad_msk_2019,31564440,PMC6949382,10.1016/j.cmet.2019.08.024,ok,
+prad_mskcc_cheny1_organoids_2014,25201530,PMC4237931,10.1016/j.cell.2014.08.016,ok,
+prad_mcspc_mskcc_2020,32220891,PMC7334067,10.1158/1078-0432.CCR-20-0168,ok,
+prad_msk_stopsack_2021,34667026,PMC8776579,10.1158/1078-0432.CCR-21-2577,ok,
+prostate_pcbm_swiss_2019,35504881,PMC9065149,10.1038/s41467-022-30003-5,ok,
+sclc_clcgp,22941188,PMC4915822,10.1038/ng.2396,ok,
+sclc_jhu,22941189,PMC3557461,10.1038/ng.2405,ok,
+skcm_broad,22817889,PMC3600117,10.1016/j.cell.2012.06.024,ok,
+rms_nih_2014,24436047,PMC4462130,10.1158/2159-8290.CD-13-0639,ok,
+sarc_tcga_pub,29100075,PMC5693358,10.1016/j.cell.2017.10.014,ok,
+sclc_cancercell_gardner_2017,28196596,PMC5313262,10.1016/j.ccell.2017.01.006,ok,
+sclc_ucologne_2015,26168399,PMC4861069,10.1038/nature14664,ok,
+sarcoma_mskcc_2022,35705560,PMC9200818,10.1038/s41467-022-30453-x,ok,
+skcm_broad_dfarber,22622578,PMC3367798,10.1038/nature11071,ok,
+skcm_yale,22842228,PMC3432702,10.1038/ng.2359,ok,
+stad_pfizer_uhongkong,24816253,PMC12468435,,error;elink,Identifier not found in PMC
+skcm_broad_brafresist_2012,24265153,PMC3947264,10.1158/2159-8290.CD-13-0617,ok,
+skcm_mskcc_2014,25409260,PMC4315319,10.1056/NEJMoa1406498,ok,
+skcm_tcga_pub_2015,26091043,PMC4580370,10.1016/j.cell.2015.05.044,ok,
+skcm_dfci_2015,26359337,PMC5054517,10.1126/science.aad0095,ok,
+stad_uhongkong,22037554,PMC12489181,,error;elink,Identifier not found in PMC
+stad_utokyo,24816255,PMC12523349,,error;elink,Identifier not found in PMC
+tet_nci_2014,24974848,PMC5705185,10.1038/ng.3016,ok,
+thyroid_mskcc_2016,26878173,PMC4767360,10.1172/JCI85271,ok,
+stes_tcga_pub,28052061,PMC5651175,10.1038/nature20805,ok,
+summit_2018,29420467,PMC5808581,10.1038/nature25475,ok,
+stmyec_wcm_2022,36577525,PMC9808553,10.1101/mcs.a006227,ok,
+ucs_jhu_2014,25233892,PMC4354107,10.1038/ncomms6006,ok,
+ucec_tcga_pub,23636398,PMC3704730,10.1038/nature12113,ok,
+um_qimr_2016,26683228,PMC4826231,10.18632/oncotarget.6614,ok,
+ucec_msk_2018,30068706,PMC6279519,10.1158/1078-0432.CCR-18-0412,ok,
+uccc_nih_2017,28485815,PMC5587124,10.1002/cncr.30745,ok,
+tmb_mskcc_2018,30643254,PMC6365097,10.1038/s41588-018-0312-8,ok,
+ucec_cptac_2020,32059776,PMC7233456,10.1016/j.cell.2020.01.026,ok,
+ucec_ccr_cfdna_msk_2022,36007103,PMC9852004,10.1158/1078-0432.CCR-22-1134,ok,
+vsc_cuk_2018,29422544,PMC5903820,10.1038/emm.2017.265,ok,
+utuc_cornell_baylor_mdacc_2019,31278255,PMC6611775,10.1038/s41467-019-10873-y,ok,
+usarc_msk_2020,32299819,PMC7367750,10.1158/1078-0432.CCR-19-3959,ok,
+utuc_igbmc_2021,33397444,PMC7780630,10.1186/s13059-020-02230-w,ok,
+lgg_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+lgg_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+lgg_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+lgg_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+lgg_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+lgg_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+lgg_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+lgg_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+lgg_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+lgg_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+lgg_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+lgg_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+crc_orion_2024,39386479,PMC11463659,10.1101/2024.09.24.614701,ok,
+brca_aurora_2023,36585450,PMC9886551,10.1038/s43018-022-00491-x,ok,
+schw_ctf_synodos_2025,33025139,PMC7785562,10.1007/s00401-020-02230-x,ok,
+ovary_geomx_gray_foundation_2024,39386723,PMC11463462,10.1101/2024.09.25.615007,ok,
+brca_tcga_pub2015,26451490,PMC4603750,10.1016/j.cell.2015.09.033,ok,
+hnsc_tcga_pub,25631445,PMC4311405,10.1038/nature14129,ok,
+luad_tcga_pub,25079552,PMC4231481,10.1038/nature13385,ok,
+thca_tcga_pub,25417114,PMC4243044,10.1016/j.cell.2014.09.050,ok,
+blca_tcga_pub,24476821,PMC3962515,10.1038/nature12965,ok,
+msk_ch_2020,33106634,PMC7891089,10.1038/s41588-020-00710-0,ok,
+msk_spectrum_tme_2022,36517593,PMC9771812,10.1038/s41586-022-05496-1,ok,
+pancan_mimsi_msk_2024,39746944,PMC11696176,10.1038/s41467-024-54970-z,ok,
+mel_iatlas_riaz_nivolumab_2017,29033130,PMC5685550,10.1016/j.cell.2017.09.028,ok,
+stad_oncosg_2018,29670109,PMC5906695,10.1038/s41467-018-03828-2,ok,
+gbm_tcga_pub,18772890,PMC2671642,10.1038/nature07385,ok,
+gbm_tcga_pub2013,24120142,PMC3910500,10.1016/j.cell.2013.09.034,ok,
+odg_msk_2017,28472509,PMC5596171,10.1093/neuonc/nox086,ok,
+gbm_tcga_pan_can_atlas_2018,29625048,PMC5957518,10.1016/j.cell.2018.03.022,ok,
+gbm_tcga_pan_can_atlas_2018,29596782,PMC6075717,10.1016/j.cels.2018.03.002,ok,
+gbm_tcga_pan_can_atlas_2018,29622463,PMC6028190,10.1016/j.ccell.2018.03.007,ok,
+gbm_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+gbm_tcga_pan_can_atlas_2018,29625055,PMC6066282,10.1016/j.cell.2018.02.052,ok,
+gbm_tcga_pan_can_atlas_2018,29625050,PMC6070353,10.1016/j.cell.2018.03.035,ok,
+gbm_tcga_pan_can_atlas_2018,29617662,PMC5916809,10.1016/j.celrep.2018.03.050,ok,
+gbm_tcga_pan_can_atlas_2018,30643250,PMC12521747,,error;elink,Identifier not found in PMC
+gbm_tcga_pan_can_atlas_2018,32214244,PMC7500457,10.1038/s41586-020-2095-1,ok,
+gbm_tcga_pan_can_atlas_2018,29625049,PMC5916814,10.1016/j.cell.2018.03.033,ok,
+gbm_tcga_pan_can_atlas_2018,29850653,PMC5972025,10.1200/PO.17.00073,ok,
+gbm_tcga_pan_can_atlas_2018,36334560,PMC12390932,,error;elink,Identifier not found in PMC
+gbm_mayo_pdx_sarkaria_2019,31852831,PMC7056576,10.1158/1078-0432.CCR-19-0909,ok,
+gbm_columbia_2019,30742119,PMC6810613,10.1038/s41591-019-0349-y,ok,
+gbm_cptac_2021,33577785,PMC8044053,10.1016/j.ccell.2021.01.006,ok,
+msk_impact_2017,28481359,PMC5461196,10.1038/nm.4333,ok,

pull_pdfs.py

@@ -0,0 +1,55 @@
+#!/usr/bin/env python3
+import csv, sys, time, requests
+
+BASE = "https://www.cbioportal.org/api"
+HEADERS = {"Accept": "application/json"}  # add 'X-API-KEY' here if your instance needs it
+
+def get_all_studies(page_size=500):
+    # cBioPortal API supports paging via pageSize/pageNumber
+    studies = []
+    page = 0
+    while True:
+        params = {"pageSize": page_size, "pageNumber": page}
+        r = requests.get(f"{BASE}/studies", headers=HEADERS, params=params, timeout=60)
+        r.raise_for_status()
+        batch = r.json()
+        if not batch:
+            break
+        studies.extend(batch)
+        page += 1
+        # friendly throttle
+        time.sleep(0.2)
+    return studies
+
+def to_list(x):
+    if x is None:
+        return []
+    if isinstance(x, list):
+        return x
+    # some portals store comma-separated string
+    return [s.strip() for s in str(x).split(",") if s.strip()]
+
+def main(out_csv="cbioportal_study_pmids.csv"):
+    studies = get_all_studies()
+    # fields commonly present: studyId, name, shortName, cancerTypeId, description, citation, pmid, etc.
+    rows = []
+    for s in studies:
+        pmids = to_list(s.get("pmid"))
+        for pmid in pmids:
+            rows.append({
+                "studyId": s.get("studyId"),
+                #"name": s.get("name"),
+                #"pmids": ";".join(pmids) if pmids else ""
+                "pmid": pmid
+            })
+    # write CSV
+    with open(out_csv, "w", newline="", encoding="utf-8") as f:
+        #w = csv.DictWriter(f, fieldnames=["studyId", "name", "pmids"])
+        w = csv.DictWriter(f, fieldnames=["studyId", "pmids"])
+        w.writeheader()
+        w.writerows(rows)
+    print(f"wrote {len(rows)} rows to {out_csv}")
+
+if __name__ == "__main__":
+    out = sys.argv[1] if len(sys.argv) > 1 else "cbioportal_study_pmids.csv"
+    main(out)

requirements.txt

@@ -0,0 +1,20 @@
+gradio
+langchain
+sqlalchemy
+langchain-community
+langchain-openai
+pypdf
+tiktoken
+openai
+langchain-text-splitters
+pdfplumber
+pillow
+sentence-transformers
+faiss-cpu
+spacy
+tqdm
+fastmcp
+
+# Only needed if you use --ocr
+pdf2image
+pytesseract

unfetched_pmcids.tsv

@@ -0,0 +1,2 @@
+PMC12088707	PMC12088707
+PMC12404184	PMC12404184