Examples with Gene ubic.gemma.model.genome.Gene used on opensource projects

Example 26 with Gene

use of ubic.gemma.model.genome.Gene in project Gemma by PavlidisLab.

the class DatabaseViewGeneratorImpl method formatDiffExResult.

private String formatDiffExResult(ExpressionExperiment ee, DifferentialExpressionAnalysisResult probeAnalysisResult, String factorName, String factorURI, String baselineDescription) {
    CompositeSequence cs = probeAnalysisResult.getProbe();
    Collection<Gene> genes = compositeSequenceService.getGenes(cs);
    if (genes.isEmpty() || genes.size() > 1) {
        return null;
    }
    Gene g = genes.iterator().next();
    if (g.getNcbiGeneId() == null)
        return null;
    Collection<ContrastResult> contrasts = probeAnalysisResult.getContrasts();
    StringBuilder buf = new StringBuilder();
    for (ContrastResult cr : contrasts) {
        FactorValue factorValue = cr.getFactorValue();
        String direction = cr.getLogFoldChange() < 0 ? "-" : "+";
        String factorValueDescription = ExperimentalDesignUtils.prettyString(factorValue);
        buf.append(String.format("%d\t%s\t%s\t%d\t%s\t%s\t%s\t%s\t%s\n", ee.getId(), ee.getShortName(), g.getNcbiGeneId().toString(), g.getId(), factorName, factorURI, baselineDescription, factorValueDescription, direction));
    }
    return buf.toString();
}

Example 27 with Gene

use of ubic.gemma.model.genome.Gene in project Gemma by PavlidisLab.

the class ArrayDesignAnnotationServiceImpl method populateProbeNameToIdMap.

private static void populateProbeNameToIdMap(ArrayDesign arrayDesign, Map<Long, Collection<Gene>> results, Map<String, Long> probeNameToId) {
    for (CompositeSequence cs : arrayDesign.getCompositeSequences()) {
        results.put(cs.getId(), new HashSet<Gene>());
        if (probeNameToId.containsKey(cs.getName())) {
            ArrayDesignAnnotationServiceImpl.log.warn("Duplicate probe name: " + cs.getName());
        }
        probeNameToId.put(cs.getName(), cs.getId());
    }
}

Example 28 with Gene

use of ubic.gemma.model.genome.Gene in project Gemma by PavlidisLab.

the class ArrayDesignAnnotationServiceImpl method generateAnnotationFile.

@Override
public int generateAnnotationFile(Writer writer, Map<CompositeSequence, Collection<BioSequence2GeneProduct>> genesWithSpecificity, OutputType ty) throws IOException {
    int compositeSequencesProcessed = 0;
    int simple = 0;
    int empty = 0;
    int complex = 0;
    // we used LinkedHasSets to keep everything in a predictable order - this is important for the gene symbols,
    // descriptions and NCBIIds (but not important for GO terms). When a probe maps to multiple genes, we list those
    // three items for the genes in the same order. There is a feature request to make
    // the order deterministic (i.e.,lexicographic sort), this could be done by using little gene objects or whatever.
    Collection<OntologyTerm> goTerms = new LinkedHashSet<>();
    Set<String> genes = new LinkedHashSet<>();
    Set<String> geneDescriptions = new LinkedHashSet<>();
    Set<String> geneIds = new LinkedHashSet<>();
    Set<String> ncbiIds = new LinkedHashSet<>();
    Map<Gene, Collection<VocabCharacteristic>> goMappings = this.getGOMappings(genesWithSpecificity);
    for (CompositeSequence cs : genesWithSpecificity.keySet()) {
        Collection<BioSequence2GeneProduct> geneclusters = genesWithSpecificity.get(cs);
        if (++compositeSequencesProcessed % 2000 == 0 && ArrayDesignAnnotationServiceImpl.log.isInfoEnabled()) {
            ArrayDesignAnnotationServiceImpl.log.info("Processed " + compositeSequencesProcessed + "/" + genesWithSpecificity.size() + " compositeSequences " + empty + " empty; " + simple + " simple; " + complex + " complex;");
        }
        if (geneclusters.isEmpty()) {
            this.writeAnnotationLine(writer, cs.getName(), "", "", null, "", "");
            empty++;
            continue;
        }
        if (geneclusters.size() == 1) {
            // common case, do it quickly.
            BioSequence2GeneProduct b2g = geneclusters.iterator().next();
            Gene g = b2g.getGeneProduct().getGene();
            goTerms = this.getGoTerms(goMappings.get(g), ty);
            String gemmaId = g.getId() == null ? "" : g.getId().toString();
            String ncbiId = g.getNcbiGeneId() == null ? "" : g.getNcbiGeneId().toString();
            this.writeAnnotationLine(writer, cs.getName(), g.getOfficialSymbol(), g.getOfficialName(), goTerms, gemmaId, ncbiId);
            simple++;
            continue;
        }
        goTerms.clear();
        genes.clear();
        geneDescriptions.clear();
        geneIds.clear();
        ncbiIds.clear();
        for (BioSequence2GeneProduct bioSequence2GeneProduct : geneclusters) {
            Gene g = bioSequence2GeneProduct.getGeneProduct().getGene();
            genes.add(g.getOfficialSymbol());
            geneDescriptions.add(g.getOfficialName());
            geneIds.add(g.getId().toString());
            Integer ncbiGeneId = g.getNcbiGeneId();
            if (ncbiGeneId != null) {
                ncbiIds.add(ncbiGeneId.toString());
            }
            goTerms.addAll(this.getGoTerms(goMappings.get(g), ty));
        }
        String geneString = StringUtils.join(genes, "|");
        String geneDescriptionString = StringUtils.join(geneDescriptions, "|");
        String geneIdsString = StringUtils.join(geneIds, "|");
        String ncbiIdsString = StringUtils.join(ncbiIds, "|");
        this.writeAnnotationLine(writer, cs.getName(), geneString, geneDescriptionString, goTerms, geneIdsString, ncbiIdsString);
        complex++;
    }
    writer.close();
    return compositeSequencesProcessed;
}

Example 29 with Gene

use of ubic.gemma.model.genome.Gene in project Gemma by PavlidisLab.

the class GeneMultifunctionalityPopulationServiceImpl method computeMultifunctionality.

/**
 * Implementation of multifunctionality computations as described in Gillis and Pavlidis (2011) PLoS ONE 6:2:e17258.
 *
 * @param gomap gomap
 * @return map
 */
private Map<Gene, Multifunctionality> computeMultifunctionality(Map<Gene, Collection<String>> gomap) {
    assert !gomap.isEmpty();
    Map<String, Integer> goGroupSizes = new HashMap<>();
    for (Gene g : gomap.keySet()) {
        for (String go : gomap.get(g)) {
            if (!goGroupSizes.containsKey(go)) {
                goGroupSizes.put(go, 1);
            } else {
                goGroupSizes.put(go, goGroupSizes.get(go) + 1);
            }
        }
    }
    GeneMultifunctionalityPopulationServiceImpl.log.info("Computed GO group sizes");
    Map<Gene, Double> geneMultifunctionalityScore = new HashMap<>();
    Map<Gene, Multifunctionality> geneMultifunctionality = new HashMap<>();
    int numGenes = gomap.size();
    for (Gene gene : gomap.keySet()) {
        Multifunctionality mf = Multifunctionality.Factory.newInstance();
        double mfscore = 0.0;
        Collection<String> sets = gomap.get(gene);
        for (String goset : sets) {
            assert goGroupSizes.containsKey(goset);
            int inGroup = goGroupSizes.get(goset);
            assert inGroup > 0;
            int outGroup = numGenes - inGroup;
            if (outGroup == 0) {
                // this doesn't meaningfully contribute to multifunctionality since every gene has it.
                continue;
            }
            mfscore += 1.0 / (inGroup * outGroup);
        }
        assert mfscore >= 0.0 && mfscore <= 1.0;
        mf.setNumGoTerms(gomap.get(gene).size());
        mf.setScore(mfscore);
        geneMultifunctionalityScore.put(gene, mfscore);
        geneMultifunctionality.put(gene, mf);
    }
    Map<Gene, Double> rawGeneMultifunctionalityRanks = Rank.rankTransform(geneMultifunctionalityScore, true);
    assert numGenes == rawGeneMultifunctionalityRanks.size();
    for (Gene gene : rawGeneMultifunctionalityRanks.keySet()) {
        // 1-base the rank before calculating ratio
        double relRank = (rawGeneMultifunctionalityRanks.get(gene) + 1) / numGenes;
        assert relRank >= 0.0 && relRank <= 1.0;
        // big values are "more multifunctional".
        geneMultifunctionality.get(gene).setRank(Math.max(0.0, 1.0 - relRank));
    }
    GeneMultifunctionalityPopulationServiceImpl.log.info("Computed multifunctionality");
    return geneMultifunctionality;
}

Example 30 with Gene

use of ubic.gemma.model.genome.Gene in project Gemma by PavlidisLab.

the class GeneMultifunctionalityPopulationServiceImpl method updateMultifunctionality.

@Override
public void updateMultifunctionality(Taxon taxon) {
    Collection<Gene> genes = geneService.loadAll(taxon);
    if (genes.isEmpty()) {
        GeneMultifunctionalityPopulationServiceImpl.log.warn("No genes found for " + taxon);
        return;
    }
    Map<Gene, Collection<String>> gomap = this.fetchGoAnnotations(genes);
    Map<Gene, Multifunctionality> mfs = this.computeMultifunctionality(gomap);
    GeneMultifunctionalityPopulationServiceImpl.log.info("Saving multifunctionality for " + genes.size() + " genes");
    Collection<Gene> batch = new HashSet<>();
    int batchSize = 200;
    int i = 0;
    for (Gene g : genes) {
        batch.add(g);
        if (batch.size() == batchSize) {
            this.saveBatch(batch, mfs);
            batch.clear();
        }
        if (++i % 1000 == 0) {
            GeneMultifunctionalityPopulationServiceImpl.log.info("Updated " + i + " genes/" + genes.size());
        }
    }
    if (!batch.isEmpty()) {
        this.saveBatch(batch, mfs);
    }
    GeneMultifunctionalityPopulationServiceImpl.log.info("Done");
}