/**
* identify additional groups that are not directly attached to amino acids.
*
* @param mc {@link ModifiedCompound}.
* @param chain a {@link Chain}.
* @return a list of added groups.
*/
private void identifyAdditionalAttachments(
ModifiedCompound mc, List<Group> ligands, Map<String, Chain> mapChainIdChain) {
if (ligands.isEmpty()) {
return;
}
// TODO: should the additional groups only be allowed to the identified
// ligands or both amino acids and ligands? Currently only on ligands
// ligands to amino acid bonds for same modification of unknown category
// will be combined in mergeModComps()
// TODO: how about chain-chain links?
List<Group> identifiedGroups = new ArrayList<Group>();
for (StructureGroup num : mc.getGroups(false)) {
Group group;
try {
// String numIns = "" + num.getResidueNumber();
// if (num.getInsCode() != null) {
// numIns += num.getInsCode();
// }
ResidueNumber resNum = new ResidueNumber();
resNum.setChainId(num.getChainId());
resNum.setSeqNum(num.getResidueNumber());
resNum.setInsCode(num.getInsCode());
// group = chain.getGroupByPDB(numIns);
group = mapChainIdChain.get(num.getChainId()).getGroupByPDB(resNum);
} catch (StructureException e) {
logger.error("Exception: ", e);
// should not happen
continue;
}
identifiedGroups.add(group);
}
int start = 0;
int n = identifiedGroups.size();
while (n > start) {
for (Group group1 : ligands) {
for (int i = start; i < n; i++) {
Group group2 = identifiedGroups.get(i);
if (!identifiedGroups.contains(group1)) {
List<Atom[]> linkedAtoms =
StructureUtil.findAtomLinkages(group1, group2, false, bondLengthTolerance);
if (!linkedAtoms.isEmpty()) {
for (Atom[] atoms : linkedAtoms) {
mc.addAtomLinkage(
StructureUtil.getStructureAtomLinkage(atoms[0], false, atoms[1], false));
}
identifiedGroups.add(group1);
break;
}
}
}
}
start = n;
n = identifiedGroups.size();
}
}
private void processCrosslink1(
Map<Component, Set<Group>> mapCompGroups,
List<ModifiedCompound> modComps,
ProteinModification mod,
List<Component> components) {
// modified residue
// TODO: is this the correct logic for CROSS_LINK_1?
Set<Group> modifiedResidues = mapCompGroups.get(components.get(0));
if (modifiedResidues != null) {
for (Group residue : modifiedResidues) {
StructureGroup strucGroup = StructureUtil.getStructureGroup(residue, true);
ModifiedCompound modRes = new ModifiedCompoundImpl(mod, strucGroup);
modComps.add(modRes);
}
}
}
/**
* 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--;
}
}
}
}
/**
* Utility method to group child nodes by their names.
*
* @param parent parent node.
* @return Map from name to child nodes.
*/
private static Map<String, List<Node>> getChildNodes(Node parent) {
if (parent == null) return Collections.emptyMap();
Map<String, List<Node>> children = new HashMap<String, List<Node>>();
NodeList nodes = parent.getChildNodes();
int nNodes = nodes.getLength();
for (int i = 0; i < nNodes; i++) {
Node node = nodes.item(i);
if (node.getNodeType() != Node.ELEMENT_NODE) continue;
String name = node.getNodeName();
List<Node> namesakes = children.get(name);
if (namesakes == null) {
namesakes = new ArrayList<Node>();
children.put(name, namesakes);
}
namesakes.add(node);
}
return children;
}
/** Merge identified modified compounds if linked. */
private void mergeModComps(List<ModifiedCompound> modComps) {
TreeSet<Integer> remove = new TreeSet<Integer>();
int n = modComps.size();
for (int icurr = 1; icurr < n; icurr++) {
ModifiedCompound curr = modComps.get(icurr);
String id = curr.getModification().getId();
if (ProteinModificationRegistry.getById(id).getCategory() != ModificationCategory.UNDEFINED)
continue;
// find linked compounds that before curr
// List<Integer> merging = new ArrayList<Integer>();
int ipre = 0;
for (; ipre < icurr; ipre++) {
if (remove.contains(ipre)) continue;
ModifiedCompound pre = modComps.get(ipre);
if (!Collections.disjoint(pre.getGroups(false), curr.getGroups(false))) {
break;
}
}
if (ipre < icurr) {
ModifiedCompound mcKeep = modComps.get(ipre);
// merge modifications of the same type
if (mcKeep.getModification().getId().equals(id)) {
// merging the current one to the previous one
mcKeep.addAtomLinkages(curr.getAtomLinkages());
remove.add(icurr);
}
}
}
Iterator<Integer> it = remove.descendingIterator();
while (it.hasNext()) {
modComps.remove(it.next().intValue());
}
}
/**
* Record unidentifiable atom linkages in a chain. Only linkages between two residues or one
* residue and one ligand will be recorded.
*/
private void recordUnidentifiableAtomLinkages(
List<ModifiedCompound> modComps, List<Group> ligands) {
// first put identified linkages in a map for fast query
Set<StructureAtomLinkage> identifiedLinkages = new HashSet<StructureAtomLinkage>();
for (ModifiedCompound mc : modComps) {
identifiedLinkages.addAll(mc.getAtomLinkages());
}
// record
// cross link
int nRes = residues.size();
for (int i = 0; i < nRes - 1; i++) {
Group group1 = residues.get(i);
for (int j = i + 1; j < nRes; j++) {
Group group2 = residues.get(j);
List<Atom[]> linkages =
StructureUtil.findAtomLinkages(group1, group2, true, bondLengthTolerance);
for (Atom[] atoms : linkages) {
StructureAtomLinkage link =
StructureUtil.getStructureAtomLinkage(atoms[0], true, atoms[1], true);
unidentifiableAtomLinkages.add(link);
}
}
}
// attachment
int nLig = ligands.size();
for (int i = 0; i < nRes; i++) {
Group group1 = residues.get(i);
for (int j = 0; j < nLig; j++) {
Group group2 = ligands.get(j);
if (group1.equals(group2)) { // overlap between residues and ligands
continue;
}
List<Atom[]> linkages =
StructureUtil.findAtomLinkages(group1, group2, false, bondLengthTolerance);
for (Atom[] atoms : linkages) {
StructureAtomLinkage link =
StructureUtil.getStructureAtomLinkage(atoms[0], true, atoms[1], false);
unidentifiableAtomLinkages.add(link);
}
}
}
}
/**
* @param linkages
* @param atomLinkages
* @return true if atomLinkages satisfy the condition; false, otherwise.
*/
private boolean matchLinkages(List<ModificationLinkage> linkages, List<Atom[]> atomLinkages) {
int nLink = linkages.size();
if (nLink != atomLinkages.size()) {
return false;
}
for (int i = 0; i < nLink - 1; i++) {
ModificationLinkage link1 = linkages.get(i);
Atom[] atoms1 = atomLinkages.get(i);
for (int j = i + 1; j < nLink; j++) {
ModificationLinkage link2 = linkages.get(j);
Atom[] atoms2 = atomLinkages.get(j);
// check components
if (((link1.getIndexOfComponent1() == link2.getIndexOfComponent1())
!= (atoms1[0].getGroup().equals(atoms2[0].getGroup())))
|| ((link1.getIndexOfComponent1() == link2.getIndexOfComponent2())
!= (atoms1[0].getGroup().equals(atoms2[1].getGroup())))
|| ((link1.getIndexOfComponent2() == link2.getIndexOfComponent1())
!= (atoms1[1].getGroup().equals(atoms2[0].getGroup())))
|| ((link1.getIndexOfComponent2() == link2.getIndexOfComponent2())
!= (atoms1[1].getGroup().equals(atoms2[1].getGroup())))) {
return false;
}
// check atoms
String label11 = link1.getLabelOfAtomOnComponent1();
String label12 = link1.getLabelOfAtomOnComponent2();
String label21 = link2.getLabelOfAtomOnComponent1();
String label22 = link2.getLabelOfAtomOnComponent2();
if ((label11 != null && label21 != null && label11.equals(label21))
!= (atoms1[0].equals(atoms2[0]))
|| (label11 != null && label22 != null && label11.equals(label22))
!= (atoms1[0].equals(atoms2[1]))
|| (label12 != null && label21 != null && label12.equals(label21))
!= (atoms1[1].equals(atoms2[0]))
|| (label12 != null && label22 != null && label12.equals(label22))
!= (atoms1[1].equals(atoms2[1]))) {
return false;
}
}
}
return true;
}
/**
* Read protein modifications from XML file and register them.
*
* @param isXml {@link InputStream} of the XML file.
* @throws IOException if failed to read the XML file.
* @throws ParserConfigurationException if parse errors occur.
* @throws SAXException the {@link DocumentBuilder} cannot be created.
*/
public static void registerProteinModificationFromXml(InputStream isXml)
throws IOException, ParserConfigurationException, SAXException {
if (isXml == null) {
throw new IllegalArgumentException("Null argument.");
}
DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
DocumentBuilder builder = factory.newDocumentBuilder();
Document doc = builder.parse(isXml);
NodeList modNodes = doc.getElementsByTagName("Entry");
int modSize = modNodes.getLength();
List<Node> nodes;
for (int iMod = 0; iMod < modSize; iMod++) {
Node modNode = modNodes.item(iMod);
Map<String, List<Node>> infoNodes = getChildNodes(modNode);
// ID
nodes = infoNodes.get("Id");
if (nodes == null || nodes.size() != 1) {
throw new RuntimeException("Each modification must have exact " + "one <Id> field.");
}
String id = nodes.get(0).getTextContent();
// modification category
nodes = infoNodes.get("Category");
if (nodes == null || nodes.size() != 1) {
throw new RuntimeException(
"Each modification must have exact "
+ "one <Category> field. See Modification "
+ id
+ ".");
}
ModificationCategory cat = ModificationCategory.getByLabel(nodes.get(0).getTextContent());
if (cat == null) {
throw new RuntimeException(
nodes.get(0).getTextContent()
+ " is not defined as an modification category."
+ " See Modification "
+ id
+ ".");
}
// occurrence type
nodes = infoNodes.get("Occurrence");
if (nodes == null || nodes.size() != 1) {
throw new RuntimeException(
"Each modification must have exact "
+ "one <Occurrence> field. See Modification "
+ id
+ ".");
}
ModificationOccurrenceType occType =
ModificationOccurrenceType.getByLabel(nodes.get(0).getTextContent());
if (occType == null) {
throw new RuntimeException(
nodes.get(0).getTextContent()
+ " is not defined as an modification occurence type."
+ " See Modification "
+ id
+ ".");
}
// condition
ModificationCondition condition = null;
{
nodes = infoNodes.get("Condition");
if (nodes == null || nodes.size() != 1) {
throw new RuntimeException(
"Each modification must have exact "
+ "one <Condition> field. See Modification "
+ id
+ ".");
}
Node compsNode = nodes.get(0);
// keep track of the labels of component indices
Map<String, Integer> mapLabelComp = new HashMap<String, Integer>();
Map<String, List<Node>> compInfoNodes = getChildNodes(compsNode);
// components
List<Node> compNodes = compInfoNodes.get("Component");
int sizeComp = compNodes.size();
List<Component> comps = new ArrayList<Component>(sizeComp);
for (int iComp = 0; iComp < sizeComp; iComp++) {
Node compNode = compNodes.get(iComp);
// comp label
NamedNodeMap compNodeAttrs = compNode.getAttributes();
Node labelNode = compNodeAttrs.getNamedItem("component");
if (labelNode == null) {
throw new RuntimeException(
"Each component must have a label." + " See Modification " + id + ".");
}
String label = labelNode.getTextContent();
if (mapLabelComp.containsKey(label)) {
throw new RuntimeException(
"Each component must have a unique label." + " See Modification " + id + ".");
}
// comp PDBCC ID
Set<String> compIds = new HashSet<String>();
List<Node> compIdNodes = getChildNodes(compNode).get("Id");
if (compIdNodes != null) {
for (Node compIdNode : compIdNodes) {
NamedNodeMap compIdNodeAttr = compIdNode.getAttributes();
Node compIdSource = compIdNodeAttr.getNamedItem("source");
if (compIdSource != null && compIdSource.getTextContent().equals("PDBCC")) {
String strComps = compIdNode.getTextContent();
if (strComps.isEmpty()) {
throw new RuntimeException("Empty component." + " See Modification " + id + ".");
}
compIds.addAll(Arrays.asList(strComps.split(",")));
}
}
}
if (compIds.isEmpty()) {
throw new RuntimeException(
"Each component must have a PDBCC ID." + " See Modification " + id + ".");
}
// terminal
boolean nTerminal = false;
boolean cTerminal = false;
List<Node> compTermNode = getChildNodes(compNode).get("Terminal");
if (compTermNode != null) {
if (compTermNode.size() != 1) {
throw new RuntimeException(
"Only one <Terminal> condition is allowed for "
+ "each component. See Modification "
+ id
+ ".");
}
String nc = compTermNode.get(0).getTextContent();
if (nc.equals("N")) {
nTerminal = true;
} else if (nc.equals("C")) {
cTerminal = true;
} else {
throw new RuntimeException(
"Only N or C is allowed for <Terminal>." + " See Modification " + id + ".");
}
}
// register
Component comp = Component.of(compIds, nTerminal, cTerminal);
comps.add(comp);
mapLabelComp.put(label, comps.size() - 1);
}
// bonds
List<Node> bondNodes = compInfoNodes.get("Bond");
List<ModificationLinkage> linkages = null;
if (bondNodes != null) {
int sizeBonds = bondNodes.size();
linkages = new ArrayList<ModificationLinkage>(sizeBonds);
for (int iBond = 0; iBond < sizeBonds; iBond++) {
Node bondNode = bondNodes.get(iBond);
Map<String, List<Node>> bondChildNodes = getChildNodes(bondNode);
if (bondChildNodes == null) {
throw new RuntimeException(
"Each bond must contain two atoms" + " See Modification " + id + ".");
}
List<Node> atomNodes = bondChildNodes.get("Atom");
if (atomNodes == null || atomNodes.size() != 2) {
throw new RuntimeException(
"Each bond must contain two atoms" + " See Modification " + id + ".");
}
// atom 1
NamedNodeMap atomNodeAttrs = atomNodes.get(0).getAttributes();
Node compNode = atomNodeAttrs.getNamedItem("component");
if (compNode == null) {
throw new RuntimeException(
"Each atom must on a component." + " See Modification " + id + ".");
}
String labelComp1 = compNode.getTextContent();
int iComp1 = mapLabelComp.get(labelComp1);
Node labelNode = atomNodeAttrs.getNamedItem("atom");
String labelAtom1 = labelNode == null ? null : labelNode.getTextContent();
String atom1 = atomNodes.get(0).getTextContent();
if (atom1.isEmpty()) {
throw new RuntimeException(
"Each atom must have a name. Please use wildcard * if unknown."
+ " See Modification "
+ id
+ ".");
}
List<String> potentialAtoms1 = Arrays.asList(atom1.split(","));
// atom 2
atomNodeAttrs = atomNodes.get(1).getAttributes();
compNode = atomNodeAttrs.getNamedItem("component");
if (compNode == null) {
throw new RuntimeException(
"Each atom must on a component." + " See Modification " + id + ".");
}
String labelComp2 = compNode.getTextContent();
int iComp2 = mapLabelComp.get(labelComp2);
labelNode = atomNodeAttrs.getNamedItem("atom");
String labelAtom2 = labelNode == null ? null : labelNode.getTextContent();
String atom2 = atomNodes.get(1).getTextContent();
if (atom2.isEmpty()) {
throw new RuntimeException(
"Each atom must have a name. Please use wildcard * if unknown."
+ " See Modification "
+ id
+ ".");
}
List<String> potentialAtoms2 = Arrays.asList(atom2.split(","));
// add linkage
ModificationLinkage linkage =
new ModificationLinkage(
comps,
iComp1,
potentialAtoms1,
labelAtom1,
iComp2,
potentialAtoms2,
labelAtom2);
linkages.add(linkage);
}
}
condition = new ModificationConditionImpl(comps, linkages);
} // end of condition
ProteinModificationImpl.Builder modBuilder =
new ProteinModificationImpl.Builder(id, cat, occType, condition);
// description
nodes = infoNodes.get("Description");
if (nodes != null && !nodes.isEmpty()) {
modBuilder.setDescription(nodes.get(0).getTextContent());
}
// cross references
nodes = infoNodes.get("CrossReference");
if (nodes != null) {
for (Node node : nodes) {
Map<String, List<Node>> xrefInfoNodes = getChildNodes(node);
// source
List<Node> xrefNode = xrefInfoNodes.get("Source");
if (xrefNode == null || xrefNode.size() != 1) {
throw new RuntimeException(
"Error in XML file: "
+ "a cross reference must contain exactly one <Source> field."
+ " See Modification "
+ id
+ ".");
}
String xrefDb = xrefNode.get(0).getTextContent();
// id
xrefNode = xrefInfoNodes.get("Id");
if (xrefNode == null || xrefNode.size() != 1) {
throw new RuntimeException(
"Error in XML file: "
+ "a cross reference must contain exactly one <Id> field."
+ " See Modification "
+ id
+ ".");
}
String xrefId = xrefNode.get(0).getTextContent();
// name
String xrefName = null;
xrefNode = xrefInfoNodes.get("Name");
if (xrefNode != null && !xrefNode.isEmpty()) {
xrefName = xrefNode.get(0).getTextContent();
}
if (xrefDb.equals("PDBCC")) {
modBuilder.setPdbccId(xrefId).setPdbccName(xrefName);
} else if (xrefDb.equals("RESID")) {
modBuilder.setResidId(xrefId).setResidName(xrefName);
} else if (xrefDb.equals("PSI-MOD")) {
modBuilder.setPsimodId(xrefId).setPsimodName(xrefName);
}
}
} // end of cross references
// formula
nodes = infoNodes.get("Formula");
if (nodes != null && !nodes.isEmpty()) {
modBuilder.setFormula(nodes.get(0).getTextContent());
}
// keywords
nodes = infoNodes.get("Keyword");
if (nodes != null && !nodes.isEmpty()) {
for (Node node : nodes) {
modBuilder.addKeyword(node.getTextContent());
}
}
ProteinModificationRegistry.register(modBuilder.build());
}
}
/** 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);
}
}
