From 510baa761f331dc3f35f075f5fa102cb42f88ffe Mon Sep 17 00:00:00 2001 From: Default user Date: Thu, 19 Oct 2023 12:55:06 +0000 Subject: [PATCH] Fix resolve smiles error --- .gitignore | 3 ++- src/train.py | 17 ++++++++++------- 2 files changed, 12 insertions(+), 8 deletions(-) diff --git a/.gitignore b/.gitignore index dc19edb..d9c9d57 100644 --- a/.gitignore +++ b/.gitignore @@ -8,4 +8,5 @@ MolecularTransformerEmbeddings/ tmp/ .venv __pycache__/ -.pytest_cache \ No newline at end of file +.pytest_cache +.ipynb_checkpoints/ diff --git a/src/train.py b/src/train.py index ca10e91..3b3d52f 100644 --- a/src/train.py +++ b/src/train.py @@ -263,15 +263,17 @@ def train(): # failed_conversion.append(row['drug']) # continue drug_id = pubchem_ids[f"DRUGBANK:{row['drug']}"] - if not drug_id or not drug_id.lower().startswith("pubchem.compound:"): - failed_conversion.append(f"{row['drug']} > {drug_id}") - continue # pubchem = normalize_id_to_translator() - drug_smiles, drug_label = get_smiles_for_drug(drug_id) - vector_list.append({"vector": vector, "payload": {"id": drug_id, "sequence": drug_smiles, "label": drug_label}}) - + try: + drug_smiles, drug_label = get_smiles_for_drug(drug_id) + vector_list.append({"vector": vector, "payload": {"id": drug_id, "sequence": drug_smiles, "label": drug_label}}) + except: + failed_conversion.append(drug_id) + + log.info(f"⚠️ Failed to get SMILES for {len(failed_conversion)} drugs：{failed_conversion}") vectordb.add("drug", vector_list) + print(f"{len(failed_conversion)} drugs ignored:") print("\n".join(failed_conversion)) @@ -287,7 +289,8 @@ def train(): # nb_model = GaussianNB() # lr_model = linear_model.LogisticRegression() - rf_model = ensemble.RandomForestClassifier(n_estimators=200, n_jobs=-1) + # rf_model = ensemble.RandomForestClassifier(n_estimators=200, n_jobs=-1) + rf_model = ensemble.RandomForestClassifier(n_estimators=200, criterion='log_loss', max_depth=None, min_samples_split=2, min_samples_leaf=1, max_features='sqrt', n_jobs=-1) # clfs = [('Naive Bayes',nb_model),('Logistic Regression',lr_model),('Random Forest',rf_model)] clfs = [("Random Forest", rf_model)]