Merge pull request #4 from compomics/cache-residue-encoding

Add caching for encoding individual residues in PeptideGraphEncoder

.gitignore

@@ -1,3 +1,4 @@
+lightning_logs/
 notebooks/_*.ipynb
 
 # vscode

molexpress/datasets/encoders.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
-from typing import Union
+from functools import lru_cache
+from typing import Dict, Tuple, Union
 
 import numpy as np
 
@@ -23,18 +24,28 @@ def __call__(self, residues: list[types.Molecule | types.SMILES | types.InChI])
         residue_graphs = []
         residue_sizes = []
         for residue in residues:
-            residue = chem_ops.get_molecule(residue)
-            residue_graph = {
-                **self.node_encoder(residue),
-                **self.edge_encoder(residue)
-            }
+            residue_graph, residue_size = self._encode_residue(
+                residue, self.node_encoder, self.edge_encoder
+            )
             residue_graphs.append(residue_graph)
-            residue_sizes.append(residue.GetNumAtoms())
+            residue_sizes.append(residue_size)
+
         disjoint_peptide_graph = self._merge_molecular_graphs(residue_graphs)
         disjoint_peptide_graph["residue_size"] = np.array(residue_sizes)
         disjoint_peptide_graph["peptide_size"] = np.array([len(residues)], dtype="int32")
         return disjoint_peptide_graph
 
+    @staticmethod
+    @lru_cache(maxsize=None)
+    def _encode_residue(
+        residue: types.Molecule | types.SMILES | types.InChI,
+        node_encoder: MolecularNodeEncoder,
+        edge_encoder: MolecularEdgeEncoder,
+    ) -> Tuple[Dict[str, np.ndarray], int]:
+        residue = chem_ops.get_molecule(residue)
+        residue_graph = {**node_encoder(residue), **edge_encoder(residue)}
+        return residue_graph, residue.GetNumAtoms()
+
     @staticmethod
     def collate_fn(
         data: list[Union[types.MolecularGraph, tuple[types.MolecularGraph, np.ndarray]]],
@@ -55,12 +66,12 @@ def collate_fn(
         disjoint_peptide_batch_graph = PeptideGraphEncoder._merge_molecular_graphs(
             disjoint_peptide_graphs
         )
-        disjoint_peptide_batch_graph["peptide_size"] = np.concatenate([
-            g["residue_size"].shape[:1] for g in disjoint_peptide_graphs
-        ]).astype("int32")
-        disjoint_peptide_batch_graph["residue_size"] = np.concatenate([
-            g["residue_size"] for g in disjoint_peptide_graphs
-        ]).astype("int32")
+        disjoint_peptide_batch_graph["peptide_size"] = np.concatenate(
+            [g["residue_size"].shape[:1] for g in disjoint_peptide_graphs]
+        ).astype("int32")
+        disjoint_peptide_batch_graph["residue_size"] = np.concatenate(
+            [g["residue_size"] for g in disjoint_peptide_graphs]
+        ).astype("int32")
 
         if y is None:
             return disjoint_peptide_batch_graph
@@ -71,21 +82,18 @@ def collate_fn(
     def _merge_molecular_graphs(
         molecular_graphs: list[types.MolecularGraph],
     ) -> types.MolecularGraph:
-
-        num_nodes = np.array([
-            g["node_state"].shape[0] for g in molecular_graphs
-        ])
+        num_nodes = np.array([g["node_state"].shape[0] for g in molecular_graphs])
 
         disjoint_molecular_graph = {}
 
-        disjoint_molecular_graph["node_state"] = np.concatenate([
-            g["node_state"] for g in molecular_graphs
-        ])
+        disjoint_molecular_graph["node_state"] = np.concatenate(
+            [g["node_state"] for g in molecular_graphs]
+        )
 
         if "edge_state" in molecular_graphs[0]:
-            disjoint_molecular_graph["edge_state"] = np.concatenate([
-                g["edge_state"] for g in molecular_graphs
-            ])
+            disjoint_molecular_graph["edge_state"] = np.concatenate(
+                [g["edge_state"] for g in molecular_graphs]
+            )
 
         edge_src = np.concatenate([graph["edge_src"] for graph in molecular_graphs])
         edge_dst = np.concatenate([graph["edge_dst"] for graph in molecular_graphs])
@@ -147,7 +155,7 @@ def __call__(self, molecule: types.Molecule) -> np.ndarray:
         if molecule.GetNumBonds() == 0:
             edge_state = np.zeros(
                 shape=(int(self.self_loops), self.output_dim + int(self.self_loops)),
-                dtype=self.output_dtype
+                dtype=self.output_dtype,
             )
             return {
                 "edge_src": edge_src,

molexpress/layers/gin_conv.py

@@ -126,7 +126,7 @@ def call(self, inputs: types.MolecularGraph) -> types.MolecularGraph:
             node_state_updated = self.activation(node_state_updated)
 
         if self.skip_connection:
-            node_state_updated += node_state
+            node_state_updated = node_state_updated + node_state
 
         if self.dropout_rate:
             node_state_updated = self.dropout(node_state_updated)