private void writeBlastResults(
Writer writer, BlastResults blastResult, FastaSequence fastaSequence) throws IOException {
// candidate uniprot
writer.write(NEW_COLUMN);
writer.write(blastResult.getAccession());
// write identity
writer.write(NEW_COLUMN);
writer.write(Float.toString(blastResult.getIdentity()));
// write sequence coverages
writer.write(NEW_COLUMN);
writer.write(
Float.toString(getQueryCoveragePercentFor(blastResult, fastaSequence.getSequence())));
writer.write(NEW_COLUMN);
writer.write(Float.toString(getMatchCoveragePercentFor(blastResult)));
// write start/end
writer.write(NEW_COLUMN);
writer.write(Integer.toString(blastResult.getStartQuery()));
writer.write("-");
writer.write(Integer.toString(blastResult.getEndQuery()));
writer.write(NEW_COLUMN);
writer.write(Integer.toString(blastResult.getStartMatch()));
writer.write("-");
writer.write(Integer.toString(blastResult.getEndMatch()));
// write query sequence
writer.write(NEW_COLUMN);
writer.write(fastaSequence.getSequence());
// write alignment
writer.write(NEW_COLUMN);
writer.write(blastResult.getAlignment());
// write matching sequence
writer.write(NEW_COLUMN);
writer.write(blastResult.getSequence());
writer.write(NEW_LINE);
writer.flush();
}
/**
* Check that all the ranges of the features that the Intact entry can contain in Intact are not
* in conflict with the new sequence(s) proposed by the blast on Swissprot to replace the Trembl
* entry.
*
* @param context : the context of the protein
* @return the trembl accession in the context if the sequences of the Swissprot proteins have
* some conflicts with the ranges of some features, the Swissprot accession if there is only
* one Swissprot protein with no conflicts and null if there are several Swissprot proteins
* with no conflict.
* @throws uk.ac.ebi.intact.protein.mapping.actions.exception.ActionProcessingException
*/
public String runAction(IdentificationContext context) throws ActionProcessingException {
// always clear the list of reports from previous actions
this.listOfReports.clear();
IntactContext intactContext = IntactContext.getCurrentInstance();
// We need to have a specific context containing the previous Trembl accession which totally
// matched the Intact protein and the possible
// proteins from Swissprot which can replace the Trembl match
if (!(context instanceof FeatureRangeCheckingContext)) {
throw new ActionProcessingException(
"We can't process a feature range checking if the context is a "
+ context.getClass().getSimpleName()
+ " and not a FeatureRangeCheckingContext instance.");
} else {
FeatureRangeCheckingContext processContext = (FeatureRangeCheckingContext) context;
int initialNumberOfBlastProtein = processContext.getResultsOfSwissprotRemapping().size();
// Create a DefaultBlastReport
BlastReport<BlastResults> report =
getReportsFactory().getBlastReport(ActionName.feature_range_checking);
this.listOfReports.add(report);
// If there were no Swissprot proteins which can replace the Trembl entry, it is an error and
// this action fails
if (processContext.getResultsOfSwissprotRemapping().isEmpty()) {
Status status =
new Status(
StatusLabel.FAILED,
"We don't have any valid results from the Swissprot-remapping process, so we will keep the Trembl entry "
+ processContext.getTremblAccession());
report.setStatus(status);
return processContext.getTremblAccession();
}
DaoFactory factory = intactContext.getDaoFactory();
// get the components involving the Intact entry
List<Component> components =
factory.getComponentDao().getByInteractorAc(processContext.getIntactAccession());
// If there is no component containing this protein, we don't have conflicts with feature
// ranges
if (components.isEmpty()) {
report.getBlastMatchingProteins().addAll(processContext.getResultsOfSwissprotRemapping());
} else {
// to check that at least one component has a feature
boolean hasAtLeastOneFeature = false;
// to check when a conflict has been detected
boolean hasRangeConflict = false;
for (Component component : components) {
Collection<Feature> features = component.getBindingDomains();
if (!features.isEmpty()) {
hasAtLeastOneFeature = true;
for (Feature feature : features) {
Collection<Range> ranges = feature.getRanges();
for (Range range : ranges) {
// undetermined ranges are not affected by the new sequence
if (!range.isUndetermined()) {
for (BlastResults protein : processContext.getResultsOfSwissprotRemapping()) {
if (!checkRangeValidWithNewSequence(range, protein, report)) {
hasRangeConflict = true;
} else {
// The sequence of this Blast protein is not in conflict with the range, we
// can keep it among the Blast results in the report
report.addBlastMatchingProtein(protein);
}
}
}
}
}
}
}
// If all the components didn't contain any feature, there is no conflict, we can keep the
// previous blast results
if (!hasAtLeastOneFeature) {
report.getBlastMatchingProteins().addAll(processContext.getResultsOfSwissprotRemapping());
} else {
// there is no conflict, we can keep the previous blast results
if (!hasRangeConflict) {
report
.getBlastMatchingProteins()
.addAll(processContext.getResultsOfSwissprotRemapping());
}
}
}
if (report.getBlastMatchingProteins().isEmpty()) {
Status status =
new Status(
StatusLabel.FAILED,
"The swissprot remapping is not possible as there are some conflicts between the sequence of the Swissprot entry and some feature ranges of the protein "
+ processContext.getIntactAccession()
+ ". We will keep the Trembl entry "
+ processContext.getTremblAccession());
report.setStatus(status);
return processContext.getTremblAccession();
} else if (report.getBlastMatchingProteins().size() < initialNumberOfBlastProtein) {
Status status =
new Status(
StatusLabel.TO_BE_REVIEWED,
processContext.getResultsOfSwissprotRemapping().size()
+ " Swissprot entries on the initial "
+ initialNumberOfBlastProtein
+ " matching Swissprot proteins have a conflict between their sequence and some feature ranges of the protein "
+ processContext.getIntactAccession());
report.setStatus(status);
return processContext.getTremblAccession();
} else {
if (report.getBlastMatchingProteins().size() == 1) {
BlastResults swissprot = report.getBlastMatchingProteins().iterator().next();
Status status =
new Status(
StatusLabel.COMPLETED,
"We don't have any conflicts between the sequence of the Swissprot entry "
+ swissprot.getAccession()
+ " and the feature ranges of the protein "
+ processContext.getIntactAccession());
report.setStatus(status);
return swissprot.getAccession();
} else {
Status status =
new Status(
StatusLabel.COMPLETED,
"We don't have any conflicts between the sequence(s) of the "
+ report.getBlastMatchingProteins().size()
+ " possible Swissprot proteins and the feature ranges of the protein "
+ processContext.getIntactAccession());
report.setStatus(status);
ArrayList<BlastResults> proteins = new ArrayList<BlastResults>();
// merge the isoforms
proteins.addAll(report.getBlastMatchingProteins());
Set<String> accessions = mergeIsoformsFromBlastProteins(proteins);
if (accessions.size() == 1) {
return accessions.iterator().next();
} else {
return processContext.getTremblAccession();
}
}
}
}
}