initial port of the CyrptoCEN space

app.py

@@ -0,0 +1,138 @@
+import datasets
+import streamlit as st
+import numpy as np
+import pandas as pd
+import altair as alt
+
+st.set_page_config(layout='wide')
+
+# parse out gene_ids from URL query args to it's possible to link to this page
+query_params = st.query_params
+if "gene_id_1" in query_params.keys():
+    gene_id_1 = query_params["gene_id_1"]
+else:
+    gene_id_1 = "TGME49_231630"
+
+if "gene_id_2" in query_params.keys():
+    gene_id_2 = query_params["gene_id_2"]
+else:
+    gene_id_2 = "TGME49_230210"
+
+    
+
+
+st.markdown("""
+# ToxoCEN Expression Scatter
+**ToxoCEN** is a co-expression network for *Toxoplasma gondii* built on 719oo RNA-seq runs across 39 studies.
+A pair of genes are said to be co-expressed when their expression is correlated across different conditions and
+is often a marker for genes to be involved in similar processes. 
+
+To Cite:
+CS Arnold, Y Wang, VB Carruthers, MJ O'Meara
+ToxoCEN: A Co-Expression Network for Toxoplasma gondii
+
+* Code available at https://github.com/maomlab/CalCEN/tree/master/vignettes/ToxoCEN
+* Full network and dataset: https://huggingface.co/datasets/maomlab/ToxoCEN
+
+## Plot scatter plot expression for a pair of genes across studies.
+Put in the ``TGME49_######`` gene_id for two genes.
+""")
+
+TGME49_transcript_annotations = datasets.load_dataset(
+    path = "maomlab/ToxoCEN",
+    data_files = {"TGME49_transcript_annotations": "TGME49_transcript_annotations.tsv"})
+TGME49_transcript_annotations = TGME49_transcript_annotations["TGME49_transcript_annotations"].to_pandas()
+
+estimated_expression_meta = datasets.load_dataset(
+    path = "maomlab/ToxoCEN",
+    data_files = {"estimated_expression_meta": "data/estimated_expression_meta.tsv"})
+estimated_expression_meta = estimated_expression_meta["estimated_expression_meta"].to_pandas()
+
+estimated_expression = datasets.load_dataset(
+    path = "maomlab/ToxoCEN",
+    data_files = {"estimated_expression": "data/estimated_expression.parquet"})
+estimated_expression = estimated_expression["estimated_expression"].to_pandas()
+
+#DEBUG
+print(f"estimated_expression shape: {estimated_expression.shape}")
+
+col1, col2, col3, padding = st.columns(spec = [0.2, 0.2, 0.2, 0.4])
+with col1:
+    gene_id_1 = st.text_input(
+        label = "Gene ID 1",
+        value = f"{gene_id_1}",
+        max_chars = 10,
+        help = "TGME49 Gene ID e.g. TGME49_231630")
+
+with col2:
+    gene_id_2 = st.text_input(
+        label = "Gene ID 2",
+        value = f"{gene_id_2}",
+        max_chars = 10,
+        help = "TGME49 Gene ID e.g. TGME49_230210")
+
+# check the user input
+try:
+    TGME49_id_1 = TGME49_transcript_annotations.loc[TGME49_transcript_annotations["gene_id"] == gene_id_1]["TGME49_id"].values[0]
+    gene_name_1 = TGME49_transcript_annotations.loc[TGME49_transcript_annotations["gene_id"] == gene_id_1]["gene_name"].values[0]
+    description_1 = TGME49_transcript_annotations.loc[TGME49_transcript_annotations["gene_id"] == gene_id_1]["description"].values[0]
+except:
+    st.error(f"Unable to locate TGME49_id for Gene ID 1: {gene_id_1}, it should be of the form 'TGME49_######'")
+
+try:
+    TGME49_id_2 = TGME49_transcript_annotations.loc[TGME49_transcript_annotations["gene_id"] == gene_id_2]["TGME49_id"].values[0]
+    gene_name_2 = TGME49_transcript_annotations.loc[TGME49_transcript_annotations["gene_id"] == gene_id_2]["gene_name"].values[0]
+    description_2 = TGME49_transcript_annotations.loc[TGME49_transcript_annotations["gene_id"] == gene_id_2]["description"].values[0]
+except:
+    st.error(f"Unable to locate TGME49_id for Gene ID 2: {gene_id_2}, it should be of the form 'TGME49_######'")
+
+chart_data = pd.DataFrame({
+    "gene_id_1": gene_id_1,
+    "gene_id_2": gene_id_2,
+    "expression_1": estimated_expression.loc[TGME49_transcript_annotations["gene_id"] == gene_id_1].to_numpy()[0],
+    "expression_2": estimated_expression.loc[TGME49_transcript_annotations["gene_id"] == gene_id_2].to_numpy()[0],
+    "log_expression_1": np.log10(estimated_expression.loc[TGME49_transcript_annotations["gene_id"] == gene_id_1].to_numpy()[0] + 1),
+    "log_expression_2": np.log10(estimated_expression.loc[TGME49_transcript_annotations["gene_id"] == gene_id_2].to_numpy()[0] + 1),
+    "run_accession": estimated_expression.columns})
+chart_data = chart_data.merge(
+    right = estimated_expression_meta,
+    on = "run_accession")
+
+with col3:
+    st.text('') # help alignment with input box
+    st.download_button(
+        label = "Download data as TSV",
+        data = chart_data.to_csv(sep ='\t').encode('utf-8'),
+        file_name = f"ToxoCEN_expression_{gene_id_1}_vs_{gene_id_2}.tsv",
+        mime = "text/csv")
+
+
+st.markdown(f"""
+#### Gene 1:
+* *Gene ID*: [{gene_id_1}](https://toxodb.org/toxodb/app/record/gene/{gene_id_1})
+{'* *Gene Symbol*:' + gene_symbol_1 if gene_symbol_1 is not None else ''}
+* *Description*: {description_1}
+* *Top [Co-Expressed Partners](https://huggingface.co/spaces/maomlab/ToxoCEN-TopHits?gene_id={gene_id_1})*
+
+#### Gene 2:
+* *Gene ID*: [{gene_id_2}](https://toxodb.org/toxodb/app/record/gene/{gene_id_2})
+{'* *Gene Symbol*:' + gene_symbol_2 if gene_symbol_2 is not None else ''}
+* *Description*: {description_2}
+* *Top [Co-Expressed Partners](https://huggingface.co/spaces/maomlab/ToxoCEN-TopHits?gene_id={gene_id_2})*
+""")
+
+chart = (
+   alt.Chart(
+       chart_data,
+       width = 750,
+       height = 750)
+   .mark_circle()
+   .encode(
+       x=alt.X("log_expression_1", title=f"Log10[{gene_id_1}+1] Expression"),
+       y=alt.Y("log_expression_2", title=f"Log10[{gene_id_2}+1] Expression"),
+       color=alt.Color("study_accession", title="Study Accession"),
+       tooltip=["run_accession", "study_accession"]))
+
+st.altair_chart(
+    chart)
+