Spaces:
Running
Running
| import time | |
| import traceback | |
| import io | |
| import pandas as pd | |
| import streamlit as st | |
| # ----------------------------------------------------------------------------- | |
| # 1. PAGE CONFIGURATION & CSS | |
| # ----------------------------------------------------------------------------- | |
| st.set_page_config(page_title="AVP-Pro", layout="wide", page_icon="🧬") | |
| GITHUB_URL = "https://github.com/wendy1031/AVP-Pro" | |
| LAB_URL = "http://www.jcu-qiulab.com" | |
| st.markdown( | |
| """ | |
| <style> | |
| #MainMenu {visibility: hidden;} | |
| footer {visibility: hidden;} | |
| .header-container { | |
| text-align: center; | |
| padding-bottom: 10px; | |
| margin-bottom: 20px; | |
| } | |
| .main-title { | |
| font-size: 3.5rem; | |
| font-weight: 700; | |
| color: #2C3E50; | |
| margin-bottom: 0px; | |
| line-height: 1.2; | |
| } | |
| .sub-title { | |
| font-size: 1.5rem; | |
| color: #5D6D7E; | |
| margin-top: 5px; | |
| font-weight: 300; | |
| } | |
| .copyright-info { | |
| font-size: 1rem; | |
| color: #2E86C1; | |
| font-weight: 600; | |
| margin-top: 10px; | |
| margin-bottom: 20px; | |
| } | |
| .copyright-info a { | |
| text-decoration: none; | |
| color: #2E86C1; | |
| } | |
| .intro-box { | |
| background-color: #f0f2f6; | |
| padding: 15px; | |
| border-radius: 10px; | |
| margin-bottom: 15px; | |
| text-align: center; | |
| font-size: 1.05rem; | |
| line-height: 1.6; | |
| color: #31333F; | |
| } | |
| div.stButton > button { | |
| border-radius: 8px; | |
| font-weight: 600; | |
| height: 3em; | |
| } | |
| div.stButton > button[kind="primary"] { | |
| background-color: #2E86C1; | |
| border: none; | |
| transition: transform 0.2s; | |
| } | |
| div.stButton > button[kind="primary"]:hover { | |
| transform: scale(1.02); | |
| } | |
| .result-card { | |
| background-color: #f8f9fa; | |
| padding: 25px; | |
| border-radius: 12px; | |
| border-left: 6px solid #2E86C1; | |
| box-shadow: 0 4px 6px rgba(0,0,0,0.05); | |
| margin-top: 20px; | |
| } | |
| </style> | |
| """, | |
| unsafe_allow_html=True, | |
| ) | |
| # ----------------------------------------------------------------------------- | |
| # 2. HELPER FUNCTIONS | |
| # ----------------------------------------------------------------------------- | |
| VALID_AA = set("ACDEFGHIKLMNPQRSTVWY") | |
| def parse_fasta(raw_text: str): | |
| """ | |
| Parses FASTA format or plain text. | |
| Returns a list of tuples: [(header, sequence), ...] | |
| """ | |
| if raw_text is None or not raw_text.strip(): | |
| return [] | |
| lines = [ln.strip() for ln in raw_text.splitlines() if ln.strip()] | |
| seqs = [] | |
| # Check if FASTA format | |
| if lines[0].startswith(">"): | |
| current_header = None | |
| current_seq = [] | |
| for line in lines: | |
| if line.startswith(">"): | |
| if current_header: | |
| full_seq = "".join(current_seq).replace(" ", "").upper() | |
| if full_seq: | |
| seqs.append((current_header, full_seq)) | |
| current_header = line[1:] # Remove > | |
| current_seq = [] | |
| else: | |
| current_seq.append(line) | |
| if current_header and current_seq: | |
| full_seq = "".join(current_seq).replace(" ", "").upper() | |
| if full_seq: | |
| seqs.append((current_header, full_seq)) | |
| else: | |
| full_seq = "".join(lines).replace(" ", "").upper() | |
| if full_seq: | |
| seqs.append(("Input_Sequence", full_seq)) | |
| return seqs | |
| def simple_seq_report(seq: str) -> dict: | |
| aa_weights = { | |
| "A": 89.09, "C": 121.15, "D": 133.10, "E": 147.13, "F": 165.19, | |
| "G": 75.07, "H": 155.16, "I": 131.17, "K": 146.19, "L": 131.17, | |
| "M": 149.21, "N": 132.12, "P": 115.13, "Q": 146.15, "R": 174.20, | |
| "S": 105.09, "T": 119.12, "V": 117.15, "W": 204.23, "Y": 181.19, | |
| } | |
| hydrophobic = set(list("AILMFWVPG")) | |
| L = len(seq) | |
| counts = {aa: seq.count(aa) for aa in VALID_AA} | |
| mw = sum(counts[aa] * aa_weights[aa] for aa in VALID_AA) - (L - 1) * 18.015 if L >= 1 else 0.0 | |
| frac_hydro = sum(counts[a] for a in hydrophobic) / L if L else 0.0 | |
| pos = set(list("KRH")) | |
| neg = set(list("DE")) | |
| net_charge = (sum(counts[a] for a in pos) - sum(counts[a] for a in neg)) | |
| return {"length": L, "mw_est": mw, "frac_hydrophobic": frac_hydro, "net_charge_est": net_charge} | |
| def get_predictor(): | |
| from avp_web.engine import AVPPredictor | |
| return AVPPredictor() | |
| def read_file_content(fname): | |
| try: | |
| with open(fname, "r", encoding="utf-8") as f: | |
| return f.read() | |
| except Exception: | |
| return None | |
| # --- Session State Initialization --- | |
| if "seq_input" not in st.session_state: | |
| st.session_state["seq_input"] = "" | |
| # 新增：用于控制文件上传组件的唯一Key | |
| if "uploader_key" not in st.session_state: | |
| st.session_state["uploader_key"] = 0 | |
| # --- Callbacks --- | |
| def load_example(): | |
| st.session_state["seq_input"] = ">Sample_AVP\nSISCSRGLVCLLPRLTNESGNDRFDS" | |
| def clear_input(): | |
| # 1. 清空文本输入变量 | |
| st.session_state["seq_input"] = "" | |
| # 2. 改变Key，强制重置文件上传组件，从而清除已上传的文件 | |
| st.session_state["uploader_key"] += 1 | |
| # ----------------------------------------------------------------------------- | |
| # 3. MAIN UI LAYOUT | |
| # ----------------------------------------------------------------------------- | |
| # Header | |
| st.markdown( | |
| f""" | |
| <div class="header-container"> | |
| <div class="main-title">AVP-Pro</div> | |
| <div class="sub-title">Deep Learning for Antiviral Peptide Discovery</div> | |
| <div class="copyright-info"> | |
| © 2025 AVP-Pro Team | <a href="{LAB_URL}" target="_blank">JCU Qiu Lab</a> | |
| </div> | |
| </div> | |
| """, | |
| unsafe_allow_html=True | |
| ) | |
| # Intro | |
| st.markdown( | |
| """ | |
| <div class="intro-box"> | |
| <b>AVP-Pro</b> is a deep learning framework designed to identify antiviral peptides (AVPs) with high precision.<br> | |
| It integrates adaptive feature fusion and contrastive learning to better capture sequence dependencies. | |
| Accurate identification of AVPs is critical for accelerating novel drug development. | |
| </div> | |
| """, | |
| unsafe_allow_html=True | |
| ) | |
| # Architecture | |
| with st.expander("🧩 View AVP-Pro Architecture (Click to expand)"): | |
| try: | |
| c1, c2, c3 = st.columns([1, 8, 1]) | |
| with c2: | |
| st.image("framework.png", caption="Figure 1: The overall architecture of AVP-Pro.", use_container_width=True) | |
| except Exception: | |
| st.warning("Framework image not found.") | |
| st.write("") | |
| # --- Input Section --- | |
| st.markdown("### ⚡ Online Prediction") | |
| st.markdown("Copy and paste your protein sequence here, or upload a FASTA file for batch prediction.") | |
| col_txt, col_btn = st.columns([3, 1]) | |
| with col_txt: | |
| st.caption("Supports FASTA format or raw peptide sequence.") | |
| with col_btn: | |
| st.button("📝 Click to load Sample AVP", on_click=load_example, use_container_width=True) | |
| # Text Area | |
| txt_val = st.text_area( | |
| label="Sequence Input", | |
| value=st.session_state["seq_input"], | |
| height=150, | |
| placeholder=">Seq1\nSISCSRGLVCLLPRLTNESGNDRFDS", | |
| label_visibility="collapsed" | |
| ) | |
| if txt_val != st.session_state["seq_input"]: | |
| st.session_state["seq_input"] = txt_val | |
| # File Uploader (注意这里绑定了 key) | |
| uploaded_file = st.file_uploader( | |
| "Or upload a sequence file (.txt / .fasta) to predict:", | |
| type=["txt", "fasta"], | |
| key=str(st.session_state["uploader_key"]) # 绑定动态Key | |
| ) | |
| # 文件加载逻辑 | |
| if uploaded_file is not None: | |
| stringio = uploaded_file.getvalue().decode("utf-8") | |
| # 只有当文件内容和当前输入框不一样时才刷新，防止死循环 | |
| if stringio != st.session_state["seq_input"]: | |
| st.session_state["seq_input"] = stringio | |
| st.rerun() | |
| # Buttons | |
| b1, b2, b3 = st.columns([1, 1, 4]) | |
| with b1: | |
| run_btn = st.button("Run Prediction", type="primary", use_container_width=True) | |
| with b2: | |
| st.button("Clear Input", on_click=clear_input, use_container_width=True) | |
| # ----------------------------------------------------------------------------- | |
| # 4. PREDICTION LOGIC | |
| # ----------------------------------------------------------------------------- | |
| if run_btn: | |
| raw_input = st.session_state["seq_input"] | |
| sequences = parse_fasta(raw_input) | |
| if not sequences: | |
| st.error("⚠️ No valid sequence found. Please input a sequence or upload a file.") | |
| else: | |
| try: | |
| predictor = get_predictor() | |
| # --- SINGLE SEQUENCE MODE --- | |
| if len(sequences) == 1: | |
| header, seq = sequences[0] | |
| invalid_chars = [c for c in seq if c not in VALID_AA] | |
| if len(seq) < 5: | |
| st.warning("Sequence too short (must be >= 5 AA).") | |
| elif invalid_chars: | |
| st.error(f"Invalid characters in sequence: {sorted(set(invalid_chars))}") | |
| else: | |
| with st.spinner("Predicting..."): | |
| res = predictor.predict(seq) | |
| rep = simple_seq_report(seq) | |
| st.markdown('<div class="result-card">', unsafe_allow_html=True) | |
| st.markdown(f"#### 🧬 Prediction Result: {header}") | |
| col_res1, col_res2 = st.columns(2) | |
| with col_res1: | |
| is_avp = res["prediction"] == 1 | |
| label_text = "AVP (Antiviral Peptide)" if is_avp else "Non-AVP" | |
| label_color = "#28a745" if is_avp else "#6c757d" | |
| st.markdown(f"**Predicted Label:** <span style='color:{label_color}; font-size:1.4rem; font-weight:bold'>{label_text}</span>", unsafe_allow_html=True) | |
| st.markdown(f"**Probability:** `{res['prob_avp']:.4f}`") | |
| st.markdown(f"**Threshold:** `{res['threshold']:.3f}`") | |
| with col_res2: | |
| st.markdown("**Sequence Properties:**") | |
| st.text(f"Length : {rep['length']}") | |
| st.text(f"Est. MW : {rep['mw_est']:.2f}") | |
| st.text(f"Net Charge : {rep['net_charge_est']}") | |
| st.markdown("</div>", unsafe_allow_html=True) | |
| # --- BATCH MODE --- | |
| else: | |
| st.info(f"📂 Detected **{len(sequences)}** sequences. Starting batch prediction...") | |
| results_list = [] | |
| progress_bar = st.progress(0) | |
| for idx, (header, seq) in enumerate(sequences): | |
| progress = (idx + 1) / len(sequences) | |
| progress_bar.progress(progress, text=f"Processing {idx+1}/{len(sequences)}") | |
| clean_seq = "".join([c for c in seq if c in VALID_AA]) | |
| if len(clean_seq) < 5: | |
| results_list.append({ | |
| "Header": header, "Label": "Error", "Probability": 0, "Note": "Length < 5" | |
| }) | |
| continue | |
| res = predictor.predict(clean_seq) | |
| label = "AVP" if res["prediction"] == 1 else "Non-AVP" | |
| results_list.append({ | |
| "Header": header, | |
| "Sequence": clean_seq, | |
| "Label": label, | |
| "Probability": round(res["prob_avp"], 4), | |
| "Threshold": round(res["threshold"], 3) | |
| }) | |
| progress_bar.empty() | |
| st.success("✅ Batch prediction complete!") | |
| df = pd.DataFrame(results_list) | |
| st.dataframe(df, use_container_width=True) | |
| csv = df.to_csv(index=False).encode('utf-8') | |
| st.download_button( | |
| label="📥 Download Results as CSV", | |
| data=csv, | |
| file_name="avp_Pro_results.csv", | |
| mime="text/csv", | |
| type="primary" | |
| ) | |
| except Exception as e: | |
| st.error("An error occurred during prediction.") | |
| st.code(traceback.format_exc()) | |
| # ----------------------------------------------------------------------------- | |
| # 5. DATASET DOWNLOAD SECTION | |
| # ----------------------------------------------------------------------------- | |
| st.write("") | |
| st.write("") | |
| st.markdown("---") | |
| st.subheader("📂 Download Benchmark Datasets") | |
| st.caption("Benchmark datasets used in AVP-Pro study.") | |
| d_col1, d_col2, d_col3 = st.columns([1, 1, 2]) | |
| avp_content = read_file_content("AVP.txt") | |
| non_avp_content = read_file_content("non_AVP.txt") | |
| with d_col1: | |
| if avp_content: | |
| st.download_button("📥 Download AVP.txt", avp_content, "AVP.txt", use_container_width=True) | |
| else: | |
| st.warning("AVP.txt missing") | |
| with d_col2: | |
| if non_avp_content: | |
| st.download_button("📥 Download non_AVP.txt", non_avp_content, "non_AVP.txt", use_container_width=True) | |
| else: | |
| st.warning("non_AVP.txt missing") | |
| st.write("") | |
| st.write("") |