fix duplicated (1/r) terms

choderalab · Sep 14, 2024 · 6e70893 · 6e70893
1 parent b383625
commit 6e70893
Showing 1 changed file with 1 addition and 3 deletions.
diff --git a/modelforge/potential/processing.py b/modelforge/potential/processing.py
@@ -495,9 +495,8 @@ def forward(self, data: Dict[str, torch.Tensor]) -> Dict[str, torch.Tensor]:
         coulomb_interactions = (
             (per_atom_charge[idx_i] * per_atom_charge[idx_j])
             * chi_r
-            / pairwise_distances
         )
-
+        # 138.96 in kj/mol nm 
         # Zero out diagonal terms (self-interaction)
         mask = torch.eye(
             coulomb_interactions.size(0), device=coulomb_interactions.device
@@ -506,7 +505,6 @@ def forward(self, data: Dict[str, torch.Tensor]) -> Dict[str, torch.Tensor]:
 
         # Sum over all interactions for each atom
         coulomb_interactions_per_atom = coulomb_interactions.sum(dim=1)
-
         data["per_atom_electrostatic_energy"] = coulomb_interactions_per_atom
 
         return data