fix handling of PDB TER records

CHANGELOG.md

@@ -9,6 +9,7 @@ Note that since we don't clearly distinguish between a public and private interf
 - [Fix] Clone ``Canvas3DParams`` when creating a ``Canvas3D`` instance to prevent shared state between multiple instances
 - Add ``includeResidueTest`` option to ``alignAndSuperposeWithSIFTSMapping``
 - Add ``parentDisplay`` param for interactions representation.
+- Fix handling of PDB TER records (#549)
 
 ## [v3.16.0] - 2022-08-25
 

src/mol-model-formats/structure/common/component.ts

@@ -32,6 +32,7 @@ const DnaAtomIdsList = [
 /** Used to reduce false positives for atom name-based type guessing */
 const NonPolymerNames = new Set([
     'FMN', 'NCN', 'FNS', 'FMA', 'ATP', 'ADP', 'AMP', 'GTP', 'GDP', 'GMP', // Mononucleotides
+    'LIG'
 ]);
 
 const StandardComponents = (function () {

src/mol-model-formats/structure/pdb/atom-site.ts

@@ -39,7 +39,7 @@ export function getAtomSiteTemplate(data: string, count: number) {
     };
 }
 
-export function getAtomSite(sites: AtomSiteTemplate, hasTer: boolean): { [K in keyof mmCIF_Schema['atom_site'] | 'partial_charge']?: CifField } {
+export function getAtomSite(sites: AtomSiteTemplate, terIndices: Set<number>): { [K in keyof mmCIF_Schema['atom_site'] | 'partial_charge']?: CifField } {
     const pdbx_PDB_model_num = CifField.ofStrings(sites.pdbx_PDB_model_num);
     const auth_asym_id = CifField.ofTokens(sites.auth_asym_id);
     const auth_seq_id = CifField.ofTokens(sites.auth_seq_id);
@@ -91,7 +91,7 @@ export function getAtomSite(sites: AtomSiteTemplate, hasTer: boolean): { [K in k
         if (asymIdCounts.has(asymId)) {
             // only change the chains name if there are TER records
             // otherwise assume repeated chain name use is from interleaved chains
-            if (hasTer && asymIdChanged) {
+            if (terIndices.size > 0 && asymIdChanged || terIndices.has(i)) {
                 const asymIdCount = asymIdCounts.get(asymId)! + 1;
                 asymIdCounts.set(asymId, asymIdCount);
                 currLabelAsymId = `${asymId}_${asymIdCount}`;

src/mol-model-formats/structure/pdb/to-cif.ts

@@ -51,7 +51,7 @@ export async function pdbToMmCif(pdb: PdbFile): Promise<CifFrame> {
 
     let modelNum = 0, modelStr = '';
     let conectRange: [number, number] | undefined = undefined;
-    let hasTer = false;
+    const terIndices = new Set<number>();
 
     for (let i = 0, _i = lines.count; i < _i; i++) {
         let s = indices[2 * i], e = indices[2 * i + 1];
@@ -164,7 +164,7 @@ export async function pdbToMmCif(pdb: PdbFile): Promise<CifFrame> {
                 break;
             case 'T':
                 if (substringStartsWith(data, s, e, 'TER')) {
-                    hasTer = true;
+                    terIndices.add(atomSite.index);
                 }
         }
     }
@@ -183,7 +183,7 @@ export async function pdbToMmCif(pdb: PdbFile): Promise<CifFrame> {
         atomSite.label_entity_id[i] = entityBuilder.getEntityId(compId, moleculeType, asymIds.value(i));
     }
 
-    const atom_site = getAtomSite(atomSite, hasTer);
+    const atom_site = getAtomSite(atomSite, terIndices);
     if (!isPdbqt) delete atom_site.partial_charge;
 
     if (conectRange) {