boltz-vsynthes 1.0.3__py3-none-any.whl → 1.0.5__py3-none-any.whl

boltz/main.py

@@ -669,46 +669,50 @@ def process_inputs(
     processed_dir = out_dir / "processed"
     processed_dir.mkdir(parents=True, exist_ok=True)
     
-    # Create structure directory
+    # Create central directories
     structure_dir = processed_dir / "structures"
+    msa_dir = processed_dir / "msa"
+    records_dir = processed_dir / "records"
+    processed_msa_dir = processed_dir / "processed" / "msa"
+    processed_constraints_dir = processed_dir / "processed" / "constraints"
+    processed_templates_dir = processed_dir / "processed" / "templates"
+    processed_mols_dir = processed_dir / "processed" / "mols"
+    predictions_dir = processed_dir / "predictions"
+    
+    # Create all central directories
     structure_dir.mkdir(parents=True, exist_ok=True)
+    msa_dir.mkdir(parents=True, exist_ok=True)
+    records_dir.mkdir(parents=True, exist_ok=True)
+    processed_msa_dir.mkdir(parents=True, exist_ok=True)
+    processed_constraints_dir.mkdir(parents=True, exist_ok=True)
+    processed_templates_dir.mkdir(parents=True, exist_ok=True)
+    processed_mols_dir.mkdir(parents=True, exist_ok=True)
+    predictions_dir.mkdir(parents=True, exist_ok=True)
     
     # Process each input file in its own directory
     for input_file in data:
         # Create a subdirectory for this input file
         file_out_dir = out_dir / input_file.stem
         file_out_dir.mkdir(parents=True, exist_ok=True)
-        
-        # Check if records exist at output path
-        records_dir = file_out_dir / "processed" / "records"
-        if records_dir.exists():
-            # Load existing records
-            existing = [Record.load(p) for p in records_dir.glob("*.json")]
-            processed_ids = {record.id for record in existing}
-            
-            # Skip if already processed
-            if input_file.stem in processed_ids:
-                click.echo(f"Found existing processed input for {input_file.stem}, skipping.")
-                all_records.extend(existing)
-                continue
 
         # Create output directories for this file
-        msa_dir = file_out_dir / "msa"
-        records_dir = file_out_dir / "processed" / "records"
-        processed_msa_dir = file_out_dir / "processed" / "msa"
-        processed_constraints_dir = file_out_dir / "processed" / "constraints"
-        processed_templates_dir = file_out_dir / "processed" / "templates"
-        processed_mols_dir = file_out_dir / "processed" / "mols"
-        predictions_dir = file_out_dir / "predictions"
-
+        file_msa_dir = file_out_dir / "msa"
+        file_records_dir = file_out_dir / "processed" / "records"
+        file_processed_msa_dir = file_out_dir / "processed" / "msa"
+        file_processed_constraints_dir = file_out_dir / "processed" / "constraints"
+        file_processed_templates_dir = file_out_dir / "processed" / "templates"
+        file_processed_mols_dir = file_out_dir / "processed" / "mols"
+        file_predictions_dir = file_out_dir / "predictions"
+
+        # Create all file-specific directories
         file_out_dir.mkdir(parents=True, exist_ok=True)
-        msa_dir.mkdir(parents=True, exist_ok=True)
-        records_dir.mkdir(parents=True, exist_ok=True)
-        processed_msa_dir.mkdir(parents=True, exist_ok=True)
-        processed_constraints_dir.mkdir(parents=True, exist_ok=True)
-        processed_templates_dir.mkdir(parents=True, exist_ok=True)
-        processed_mols_dir.mkdir(parents=True, exist_ok=True)
-        predictions_dir.mkdir(parents=True, exist_ok=True)
+        file_msa_dir.mkdir(parents=True, exist_ok=True)
+        file_records_dir.mkdir(parents=True, exist_ok=True)
+        file_processed_msa_dir.mkdir(parents=True, exist_ok=True)
+        file_processed_constraints_dir.mkdir(parents=True, exist_ok=True)
+        file_processed_templates_dir.mkdir(parents=True, exist_ok=True)
+        file_processed_mols_dir.mkdir(parents=True, exist_ok=True)
+        file_predictions_dir.mkdir(parents=True, exist_ok=True)
 
         # Load CCD
         if boltz2:
@@ -721,27 +725,34 @@ def process_inputs(
         process_input_partial = partial(
             process_input,
             ccd=ccd,
-            msa_dir=msa_dir,
+            msa_dir=file_msa_dir,
             mol_dir=mol_dir,
             boltz2=boltz2,
             use_msa_server=use_msa_server,
             msa_server_url=msa_server_url,
             msa_pairing_strategy=msa_pairing_strategy,
             max_msa_seqs=max_msa_seqs,
-            processed_msa_dir=processed_msa_dir,
-            processed_constraints_dir=processed_constraints_dir,
-            processed_templates_dir=processed_templates_dir,
-            processed_mols_dir=processed_mols_dir,
+            processed_msa_dir=file_processed_msa_dir,
+            processed_constraints_dir=file_processed_constraints_dir,
+            processed_templates_dir=file_processed_templates_dir,
+            processed_mols_dir=file_processed_mols_dir,
             structure_dir=structure_dir,  # Use the central structure directory
-            records_dir=records_dir,
+            records_dir=file_records_dir,
         )
 
         # Process this input file
         click.echo(f"Processing {input_file.name}")
         process_input_partial(input_file)
 
+        # Copy MSA files to central MSA directory
+        for msa_file in file_processed_msa_dir.glob("*.npz"):
+            target_msa_file = msa_dir / msa_file.name
+            if not target_msa_file.exists():
+                import shutil
+                shutil.copy2(msa_file, target_msa_file)
+
         # Load records for this file
-        records = [Record.load(p) for p in records_dir.glob("*.json")]
+        records = [Record.load(p) for p in file_records_dir.glob("*.json")]
         all_records.extend(records)
 
     # Create combined manifest
@@ -952,11 +963,6 @@ def cli() -> None:
     is_flag=True,
     help="Whether to disable the kernels. Default False",
 )
-@click.option(
-    "--skip_structure",
-    is_flag=True,
-    help="Skip structure prediction and go straight to affinity prediction. Default is False.",
-)
 def predict(  # noqa: C901, PLR0915, PLR0912
     data: str,
     out_dir: str,
@@ -990,7 +996,6 @@ def predict(  # noqa: C901, PLR0915, PLR0912
     subsample_msa: bool = True,
     num_subsampled_msa: int = 1024,
     no_kernels: bool = False,
-    skip_structure: bool = False,
 ) -> None:
     """Run predictions with Boltz."""
     # If cpu, write a friendly warning
@@ -1151,7 +1156,7 @@ def predict(  # noqa: C901, PLR0915, PLR0912
         precision=32 if model == "boltz1" else "bf16-mixed",
     )
 
-    if filtered_manifest.records and not skip_structure:
+    if filtered_manifest.records:
         msg = f"Running structure prediction for {len(filtered_manifest.records)} input"
         msg += "s." if len(filtered_manifest.records) > 1 else "."
         click.echo(msg)
@@ -1240,14 +1245,6 @@ def predict(  # noqa: C901, PLR0915, PLR0912
         msg += "s." if len(manifest_filtered.records) > 1 else "."
         click.echo(msg)
 
-        # When skipping structure prediction, we need to ensure structure files exist
-        if skip_structure:
-            for record in manifest_filtered.records:
-                structure_path = out_dir / "predictions" / record.id / f"pre_affinity_{record.id}.npz"
-                if not structure_path.exists():
-                    msg = f"Structure file not found for {record.id}. Cannot run affinity prediction without structure."
-                    raise FileNotFoundError(msg)
-
         pred_writer = BoltzAffinityWriter(
             data_dir=processed.targets_dir,
             output_dir=out_dir / "predictions",

{boltz_vsynthes-1.0.3.dist-info → boltz_vsynthes-1.0.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: boltz-vsynthes
-Version: 1.0.3
+Version: 1.0.5
 Summary: Boltz for V-Synthes
 Requires-Python: <3.13,>=3.10
 Description-Content-Type: text/markdown

{boltz_vsynthes-1.0.3.dist-info → boltz_vsynthes-1.0.5.dist-info}/RECORD

@@ -1,5 +1,5 @@
 boltz/__init__.py,sha256=F_-so3S40iZrSZ89Ge4TS6aZqwWyZXq_H4AXGDlbA_g,187
-boltz/main.py,sha256=p-iXEzVh_CAkKZs2MJ7EojQQdIZnC9cYXr9Bax_PrU4,41377
+boltz/main.py,sha256=sF_fNSzOElFhnlUBrnRidY1Dg_dduIHl23CREMo_ICc,41374
 boltz/data/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 boltz/data/const.py,sha256=1M-88Z6HkfKY6MkNtqcj3b9P-oX9xEXluh3qM_u8dNU,26779
 boltz/data/mol.py,sha256=maOpPHEGX1VVXCIFY6pQNGF7gUBZPAfgSvuPf2QO1yc,34268
@@ -104,9 +104,9 @@ boltz/model/optim/scheduler.py,sha256=nB4jz0CZ4pR4n08LQngExL_pNycIdYI8AXVoHPnZWQ
 boltz/model/potentials/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 boltz/model/potentials/potentials.py,sha256=vev8Vjfs-ML1hyrdv_R8DynG4wSFahJ6nzPWp7CYQqw,17507
 boltz/model/potentials/schedules.py,sha256=m7XJjfuF9uTX3bR9VisXv1rvzJjxiD8PobXRpcBBu1c,968
-boltz_vsynthes-1.0.3.dist-info/licenses/LICENSE,sha256=8GZ_1eZsUeG6jdqgJJxtciWzADfgLEV4LY8sKUOsJhc,1102
-boltz_vsynthes-1.0.3.dist-info/METADATA,sha256=KQnSi_PP-_Gkz8yG3ED3-a_R4-IUlX92QZxwG8eQe8w,7171
-boltz_vsynthes-1.0.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-boltz_vsynthes-1.0.3.dist-info/entry_points.txt,sha256=n5a5I35ntu9lmyr16oZgHPFY0b0YxjiixY7m7nbMTLc,41
-boltz_vsynthes-1.0.3.dist-info/top_level.txt,sha256=MgU3Jfb-ctWm07YGMts68PMjSh9v26D0gfG3dFRmVFA,6
-boltz_vsynthes-1.0.3.dist-info/RECORD,,
+boltz_vsynthes-1.0.5.dist-info/licenses/LICENSE,sha256=8GZ_1eZsUeG6jdqgJJxtciWzADfgLEV4LY8sKUOsJhc,1102
+boltz_vsynthes-1.0.5.dist-info/METADATA,sha256=-MCCHAI1TOA1tlDaX-X6npP-HgYgJZht77XK-J9CeAI,7171
+boltz_vsynthes-1.0.5.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+boltz_vsynthes-1.0.5.dist-info/entry_points.txt,sha256=n5a5I35ntu9lmyr16oZgHPFY0b0YxjiixY7m7nbMTLc,41
+boltz_vsynthes-1.0.5.dist-info/top_level.txt,sha256=MgU3Jfb-ctWm07YGMts68PMjSh9v26D0gfG3dFRmVFA,6
+boltz_vsynthes-1.0.5.dist-info/RECORD,,