/**
* Assembly the matched linkages.
*
* @param matchedAtomsOfLinkages
* @param mod
* @param condition
* @param ret ModifiedCompound will be stored here.
*/
private void assembleLinkages(
List<List<Atom[]>> matchedAtomsOfLinkages,
ProteinModification mod,
List<ModifiedCompound> ret) {
ModificationCondition condition = mod.getCondition();
List<ModificationLinkage> modLinks = condition.getLinkages();
int nLink = matchedAtomsOfLinkages.size();
int[] indices = new int[nLink];
Set<ModifiedCompound> identifiedCompounds = new HashSet<ModifiedCompound>();
while (indices[0] < matchedAtomsOfLinkages.get(0).size()) {
List<Atom[]> atomLinkages = new ArrayList<Atom[]>(nLink);
for (int iLink = 0; iLink < nLink; iLink++) {
Atom[] atoms = matchedAtomsOfLinkages.get(iLink).get(indices[iLink]);
atomLinkages.add(atoms);
}
if (matchLinkages(modLinks, atomLinkages)) {
// matched
int n = atomLinkages.size();
List<StructureAtomLinkage> linkages = new ArrayList<StructureAtomLinkage>(n);
for (int i = 0; i < n; i++) {
Atom[] linkage = atomLinkages.get(i);
StructureAtomLinkage link =
StructureUtil.getStructureAtomLinkage(
linkage[0], residues.contains(linkage[0].getGroup()),
linkage[1], residues.contains(linkage[1].getGroup()));
linkages.add(link);
}
ModifiedCompound mc = new ModifiedCompoundImpl(mod, linkages);
if (!identifiedCompounds.contains(mc)) {
ret.add(mc);
identifiedCompounds.add(mc);
}
}
// indices++ (e.g. [0,0,1]=>[0,0,2]=>[1,2,0])
int i = nLink - 1;
while (i >= 0) {
if (i == 0 || indices[i] < matchedAtomsOfLinkages.get(i).size() - 1) {
indices[i]++;
break;
} else {
indices[i] = 0;
i--;
}
}
}
}
private void processMultiCrosslink(
Map<Component, Set<Group>> mapCompGroups,
List<ModifiedCompound> modComps,
ProteinModification mod,
ModificationCondition condition) {
// for multiple components
// find linkages first
List<List<Atom[]>> matchedAtomsOfLinkages = getMatchedAtomsOfLinkages(condition, mapCompGroups);
if (matchedAtomsOfLinkages.size() != condition.getLinkages().size()) {
return;
}
assembleLinkages(matchedAtomsOfLinkages, mod, modComps);
}
/** Get matched atoms for all linkages. */
private List<List<Atom[]>> getMatchedAtomsOfLinkages(
ModificationCondition condition, Map<Component, Set<Group>> mapCompGroups) {
List<ModificationLinkage> linkages = condition.getLinkages();
int nLink = linkages.size();
List<List<Atom[]>> matchedAtomsOfLinkages = new ArrayList<List<Atom[]>>(nLink);
for (int iLink = 0; iLink < nLink; iLink++) {
ModificationLinkage linkage = linkages.get(iLink);
Component comp1 = linkage.getComponent1();
Component comp2 = linkage.getComponent2();
// boolean isAA1 = comp1.;
// boolean isAA2 = comp2.getType()==true;
Set<Group> groups1 = mapCompGroups.get(comp1);
Set<Group> groups2 = mapCompGroups.get(comp2);
List<Atom[]> list = new ArrayList<Atom[]>();
List<String> potentialNamesOfAtomOnGroup1 = linkage.getPDBNameOfPotentialAtomsOnComponent1();
for (String name : potentialNamesOfAtomOnGroup1) {
if (name.equals("*")) {
// wildcard
potentialNamesOfAtomOnGroup1 = null; // search all atoms
break;
}
}
List<String> potentialNamesOfAtomOnGroup2 = linkage.getPDBNameOfPotentialAtomsOnComponent2();
for (String name : potentialNamesOfAtomOnGroup2) {
if (name.equals("*")) {
// wildcard
potentialNamesOfAtomOnGroup2 = null; // search all atoms
break;
}
}
for (Group g1 : groups1) {
for (Group g2 : groups2) {
if (g1.equals(g2)) {
continue;
}
// only for wildcard match of two residues
boolean ignoreNCLinkage =
potentialNamesOfAtomOnGroup1 == null
&& potentialNamesOfAtomOnGroup2 == null
&& residues.contains(g1)
&& residues.contains(g2);
Atom[] atoms =
StructureUtil.findNearestAtomLinkage(
g1,
g2,
potentialNamesOfAtomOnGroup1,
potentialNamesOfAtomOnGroup2,
ignoreNCLinkage,
bondLengthTolerance);
if (atoms != null) {
list.add(atoms);
}
}
}
if (list.isEmpty()) {
// broken linkage
break;
}
matchedAtomsOfLinkages.add(list);
}
return matchedAtomsOfLinkages;
}
/**
* @param modifications a set of {@link ProteinModification}s.
* @param residues
* @param ligands
* @param saveTo save result to
* @return map from component to list of corresponding residues in the chain.
*/
private void addModificationGroups(
final Set<ProteinModification> modifications,
final List<Group> residues,
final List<Group> ligands,
final Map<Component, Set<Group>> saveTo) {
if (residues == null || ligands == null || modifications == null) {
throw new IllegalArgumentException("Null argument(s).");
}
Map<Component, Set<Component>> mapSingleMultiComps = new HashMap<Component, Set<Component>>();
for (ProteinModification mod : modifications) {
ModificationCondition condition = mod.getCondition();
for (Component comp : condition.getComponents()) {
for (String pdbccId : comp.getPdbccIds()) {
Component single =
Component.of(Collections.singleton(pdbccId), comp.isNTerminal(), comp.isCTerminal());
Set<Component> mult = mapSingleMultiComps.get(single);
if (mult == null) {
mult = new HashSet<Component>();
mapSingleMultiComps.put(single, mult);
}
mult.add(comp);
}
}
}
{
// ligands
Set<Component> ligandsWildCard = mapSingleMultiComps.get(Component.of("*"));
for (Group group : ligands) {
String pdbccId = group.getPDBName().trim();
Set<Component> comps = mapSingleMultiComps.get(Component.of(pdbccId));
for (Component comp : unionComponentSet(ligandsWildCard, comps)) {
Set<Group> gs = saveTo.get(comp);
if (gs == null) {
gs = new LinkedHashSet<Group>();
saveTo.put(comp, gs);
}
gs.add(group);
}
}
}
{
// residues
if (residues.isEmpty()) {
return;
}
Set<Component> residuesWildCard = mapSingleMultiComps.get(Component.of("*"));
// for all residues
for (Group group : residues) {
String pdbccId = group.getPDBName().trim();
Set<Component> comps = mapSingleMultiComps.get(Component.of(pdbccId));
for (Component comp : unionComponentSet(residuesWildCard, comps)) {
Set<Group> gs = saveTo.get(comp);
if (gs == null) {
gs = new LinkedHashSet<Group>();
saveTo.put(comp, gs);
}
gs.add(group);
}
}
// for N-terminal
int nRes = residues.size();
int iRes = 0;
Group res;
do {
// for all ligands on N terminal and the first residue
res = residues.get(iRes++);
Set<Component> nTermWildCard = mapSingleMultiComps.get(Component.of("*", true, false));
Set<Component> comps = mapSingleMultiComps.get(Component.of(res.getPDBName(), true, false));
for (Component comp : unionComponentSet(nTermWildCard, comps)) {
Set<Group> gs = saveTo.get(comp);
if (gs == null) {
gs = new LinkedHashSet<Group>();
saveTo.put(comp, gs);
}
gs.add(res);
}
} while (iRes < nRes && ligands.contains(res));
// for C-terminal
iRes = residues.size() - 1;
do {
// for all ligands on C terminal and the last residue
res = residues.get(iRes--);
Set<Component> cTermWildCard = mapSingleMultiComps.get(Component.of("*", false, true));
Set<Component> comps = mapSingleMultiComps.get(Component.of(res.getPDBName(), false, true));
for (Component comp : unionComponentSet(cTermWildCard, comps)) {
Set<Group> gs = saveTo.get(comp);
if (gs == null) {
gs = new LinkedHashSet<Group>();
saveTo.put(comp, gs);
}
gs.add(res);
}
} while (iRes >= 0 && ligands.contains(res));
}
}
/**
* Identify a set of modifications in a a list of chains.
*
* @param chains query {@link Chain}s.
* @param potentialModifications query {@link ProteinModification}s.
*/
public void identify(
final List<Chain> chains, final Set<ProteinModification> potentialModifications) {
if (chains == null) {
throw new IllegalArgumentException("Null structure.");
}
if (potentialModifications == null) {
throw new IllegalArgumentException("Null potentialModifications.");
}
reset();
if (potentialModifications.isEmpty()) {
return;
}
Map<String, Chain> mapChainIdChain = new HashMap<String, Chain>(chains.size());
residues = new ArrayList<Group>();
List<Group> ligands = new ArrayList<Group>();
Map<Component, Set<Group>> mapCompGroups = new HashMap<Component, Set<Group>>();
for (Chain chain : chains) {
mapChainIdChain.put(chain.getChainID(), chain);
List<Group> ress = StructureUtil.getAminoAcids(chain);
// List<Group> ligs = chain.getAtomLigands();
List<Group> ligs = StructureTools.filterLigands(chain.getAtomGroups());
residues.addAll(ress);
residues.removeAll(ligs);
ligands.addAll(ligs);
addModificationGroups(potentialModifications, ress, ligs, mapCompGroups);
}
if (residues.isEmpty()) {
String pdbId = "?";
if (chains.size() > 0) {
Structure struc = chains.get(0).getParent();
if (struc != null) pdbId = struc.getPDBCode();
}
logger.warn(
"No amino acids found for {}. Either you did not parse the PDB file with alignSEQRES records, or this record does not contain any amino acids.",
pdbId);
}
List<ModifiedCompound> modComps = new ArrayList<ModifiedCompound>();
for (ProteinModification mod : potentialModifications) {
ModificationCondition condition = mod.getCondition();
List<Component> components = condition.getComponents();
if (!mapCompGroups.keySet().containsAll(components)) {
// not all components exist for this mod.
continue;
}
int sizeComps = components.size();
if (sizeComps == 1) {
processCrosslink1(mapCompGroups, modComps, mod, components);
} else {
processMultiCrosslink(mapCompGroups, modComps, mod, condition);
}
}
if (recordAdditionalAttachments) {
// identify additional groups that are not directly attached to amino acids.
for (ModifiedCompound mc : modComps) {
identifyAdditionalAttachments(mc, ligands, mapChainIdChain);
}
}
mergeModComps(modComps);
identifiedModifiedCompounds.addAll(modComps);
// record unidentifiable linkage
if (recordUnidentifiableModifiedCompounds) {
recordUnidentifiableAtomLinkages(modComps, ligands);
recordUnidentifiableModifiedResidues(modComps);
}
}