Examples with SnpEffPredictionParserFactory com.github.lindenb.jvarkit.util.vcf.predictions.SnpEffPredictionParserFactory used on opensource projects

Example 1 with SnpEffPredictionParserFactory

use of com.github.lindenb.jvarkit.util.vcf.predictions.SnpEffPredictionParserFactory in project jvarkit by lindenb.

the class VCFComparePredictions method doWork.

@Override
public int doWork(List<String> args) {
    PrintWriter out = null;
    SortingCollection<LineAndFile> variants = null;
    try {
        if (args.isEmpty()) {
            LOG.error("Illegal number of arguments");
            return -1;
        }
        out = super.openFileOrStdoutAsPrintWriter(super.outputFile);
        variants = SortingCollection.newInstance(LineAndFile.class, new AbstractVCFCompareBase.LineAndFileCodec(), new AbstractVCFCompareBase.LineAndFileComparator(), super.sortingCollectionArgs.getMaxRecordsInRam(), super.sortingCollectionArgs.getTmpPaths());
        variants.setDestructiveIteration(true);
        for (final String filename : args) {
            LOG.info("Reading from " + filename);
            Input input = super.put(variants, filename);
            LOG.info("end reading " + input.filename);
        }
        List<PredictionTuple> predictionTuples = new ArrayList<PredictionTuple>(super.inputs.size());
        for (AbstractVCFCompareBase.Input input : this.inputs) {
            PredictionTuple predictionTuple = new PredictionTuple();
            predictionTuple.snpEffPredictionParser = new SnpEffPredictionParserFactory(input.codecAndHeader.header).get();
            predictionTuple.vepPredictionParser = new VepPredictionParserFactory(input.codecAndHeader.header).get();
            predictionTuples.add(predictionTuple);
        }
        List<AbstractVCFCompareBase.LineAndFile> row = new ArrayList<LineAndFile>(super.inputs.size());
        CloseableIterator<LineAndFile> iter = variants.iterator();
        final Comparator<LineAndFile> posCompare = (A, B) -> A.getContigPosRef().compareTo(B.getContigPosRef());
        for (; ; ) {
            LineAndFile rec = null;
            if (iter.hasNext()) {
                rec = iter.next();
            }
            if (rec == null || (!row.isEmpty() && posCompare.compare(row.get(0), rec) != 0)) {
                if (!row.isEmpty()) {
                    boolean printed = false;
                    VariantContext ctx = row.get(0).getContext();
                    if (row.size() != this.inputs.size()) {
                        startLine(out, ctx);
                        out.println("\tDiscordant number of variants");
                        printed = true;
                    }
                    for (int i = 0; i + 1 < row.size(); ++i) {
                        Input input1 = this.inputs.get(row.get(i).fileIdx);
                        VariantContext ctx1 = row.get(i).getContext();
                        PredictionTuple predtuple1 = predictionTuples.get(row.get(i).fileIdx);
                        List<VepPrediction> vepPredictions1 = predtuple1.vepPredictionParser.getPredictions(ctx1);
                        List<SnpEffPrediction> snpEffPredictions1 = predtuple1.snpEffPredictionParser.getPredictions(ctx1);
                        Set<SequenceOntologyTree.Term> so_vep_1 = getVepSoTerms(predtuple1.vepPredictionParser, ctx1);
                        Set<SequenceOntologyTree.Term> so_snpeff_1 = getSnpEffSoTerms(predtuple1.snpEffPredictionParser, ctx1);
                        for (int j = i + 1; j < row.size(); ++j) {
                            Input input2 = this.inputs.get(row.get(j).fileIdx);
                            VariantContext ctx2 = row.get(j).getContext();
                            PredictionTuple predtuple2 = predictionTuples.get(row.get(j).fileIdx);
                            List<VepPrediction> vepPredictions2 = predtuple2.vepPredictionParser.getPredictions(ctx2);
                            List<SnpEffPrediction> snpEffPredictions2 = predtuple2.snpEffPredictionParser.getPredictions(ctx2);
                            Set<SequenceOntologyTree.Term> so_vep_2 = getVepSoTerms(predtuple2.vepPredictionParser, ctx2);
                            Set<SequenceOntologyTree.Term> so_snpeff_2 = getSnpEffSoTerms(predtuple2.snpEffPredictionParser, ctx2);
                            if (vepPredictions1.size() != vepPredictions2.size()) {
                                startLine(out, ctx);
                                out.print("\tVEP discordant transcripts count");
                                out.print("\t" + input1.filename + ":" + vepPredictions1.size());
                                out.print("\t" + input2.filename + ":" + vepPredictions2.size());
                                out.println();
                                printed = true;
                            }
                            if (snpEffPredictions1.size() != snpEffPredictions2.size()) {
                                startLine(out, ctx);
                                out.print("\tSNPEFF discordant transcripts count");
                                out.print("\t" + input1.filename + ":" + snpEffPredictions1.size());
                                out.print("\t" + input2.filename + ":" + snpEffPredictions2.size());
                                out.println();
                                printed = true;
                            }
                            if (!unshared(so_vep_1, so_vep_2).isEmpty()) {
                                startLine(out, ctx);
                                out.print("\tVEP discordant SO:terms");
                                printDiscordantSO(out, input1, so_vep_1, input2, so_vep_2);
                                printed = true;
                            }
                            if (!unshared(so_snpeff_1, so_snpeff_2).isEmpty()) {
                                startLine(out, ctx);
                                out.print("\tSNPEFF discordant SO:terms");
                                printDiscordantSO(out, input1, so_snpeff_1, input2, so_snpeff_2);
                                printed = true;
                            }
                        }
                    }
                    if (!printed) {
                        startLine(out, ctx);
                        out.println("\tPASS");
                    }
                    row.clear();
                }
                if (rec == null)
                    break;
            }
            row.add(rec);
        }
        iter.close();
        out.flush();
        out.close();
        out = null;
        return 0;
    } catch (Exception err) {
        LOG.error(err);
        return -1;
    } finally {
        CloserUtil.close(out);
        try {
            if (variants != null)
                variants.cleanup();
        } catch (Exception err) {
        }
    }
}

Example 2 with SnpEffPredictionParserFactory

use of com.github.lindenb.jvarkit.util.vcf.predictions.SnpEffPredictionParserFactory in project jvarkit by lindenb.

the class VcfTools method init.

public void init(final VCFHeader header) {
    this.header = header;
    this.snpEffPredictionParser = new SnpEffPredictionParserFactory().header(header).get();
    this.vepPredictionParser = new VepPredictionParserFactory().header(header).get();
    this.annPredictionParser = new AnnPredictionParserFactory(header).get();
}

Example 3 with SnpEffPredictionParserFactory

use of com.github.lindenb.jvarkit.util.vcf.predictions.SnpEffPredictionParserFactory in project jvarkit by lindenb.

the class VCFCompare method doWork.

@Override
public int doWork(final List<String> args) {
    if (args.isEmpty()) {
        LOG.error("VCFs missing.");
        return -1;
    }
    if (args.size() != 2) {
        System.err.println("Illegal number or arguments. Expected two VCFs");
        return -1;
    }
    PrintWriter pw = null;
    XMLStreamWriter w = null;
    InputStream in = null;
    SortingCollection<LineAndFile> variants = null;
    try {
        LineAndFileComparator varcmp = new LineAndFileComparator();
        variants = SortingCollection.newInstance(LineAndFile.class, new LineAndFileCodec(), varcmp, this.writingSortingCollection.getMaxRecordsInRam(), this.writingSortingCollection.getTmpPaths());
        variants.setDestructiveIteration(true);
        for (int i = 0; i < 2; ++i) {
            this.inputs[i] = new Input();
            this.inputs[i].codec = VCFUtils.createDefaultVCFCodec();
            this.inputs[i].filename = args.get(i);
            LOG.info("Opening " + this.inputs[i].filename);
            in = IOUtils.openURIForReading(this.inputs[i].filename);
            final LineReader lr = new SynchronousLineReader(in);
            final LineIterator li = new LineIteratorImpl(lr);
            this.inputs[i].header = (VCFHeader) this.inputs[i].codec.readActualHeader(li);
            this.inputs[i].vepPredictionParser = new VepPredictionParserFactory(this.inputs[i].header).get();
            this.inputs[i].snpEffPredictionParser = new SnpEffPredictionParserFactory(this.inputs[i].header).get();
            this.inputs[i].annPredictionParser = new AnnPredictionParserFactory(this.inputs[i].header).get();
            while (li.hasNext()) {
                LineAndFile laf = new LineAndFile();
                laf.fileIdx = i;
                laf.line = li.next();
                variants.add(laf);
            }
            LOG.info("Done Reading " + this.inputs[i].filename);
            CloserUtil.close(li);
            CloserUtil.close(lr);
            CloserUtil.close(in);
        }
        variants.doneAdding();
        LOG.info("Done Adding");
        Set<String> commonSamples = new TreeSet<String>(this.inputs[0].header.getSampleNamesInOrder());
        commonSamples.retainAll(this.inputs[1].header.getSampleNamesInOrder());
        List<Venn0> venn1List = new ArrayList<VCFCompare.Venn0>();
        venn1List.add(new Venn1("ALL"));
        venn1List.add(new Venn1("having ID") {

            @Override
            public VariantContext filter(VariantContext ctx, int fileIndex) {
                return ctx == null || !ctx.hasID() ? null : ctx;
            }
        });
        venn1List.add(new Venn1("QUAL greater 30") {

            @Override
            public VariantContext filter(VariantContext ctx, int fileIndex) {
                return ctx == null || !ctx.hasLog10PError() || ctx.getPhredScaledQual() < 30.0 ? null : ctx;
            }
        });
        for (VariantContext.Type t : VariantContext.Type.values()) {
            venn1List.add(new VennType(t));
        }
        for (SequenceOntologyTree.Term term : SequenceOntologyTree.getInstance().getTerms()) {
            venn1List.add(new VennPred("vep", term) {

                @Override
                Set<Term> terms(VariantContext ctx, int file_id) {
                    Set<Term> tt = new HashSet<SequenceOntologyTree.Term>();
                    for (VepPredictionParser.VepPrediction pred : VCFCompare.this.inputs[file_id].vepPredictionParser.getPredictions(ctx)) {
                        tt.addAll(pred.getSOTerms());
                    }
                    return tt;
                }
            });
            venn1List.add(new VennPred("SnpEff", term) {

                @Override
                Set<Term> terms(VariantContext ctx, int file_id) {
                    Set<Term> tt = new HashSet<SequenceOntologyTree.Term>();
                    for (SnpEffPredictionParser.SnpEffPrediction pred : VCFCompare.this.inputs[file_id].snpEffPredictionParser.getPredictions(ctx)) {
                        tt.addAll(pred.getSOTerms());
                    }
                    return tt;
                }
            });
            venn1List.add(new VennPred("ANN", term) {

                @Override
                Set<Term> terms(VariantContext ctx, int file_id) {
                    Set<Term> tt = new HashSet<SequenceOntologyTree.Term>();
                    for (AnnPredictionParser.AnnPrediction pred : VCFCompare.this.inputs[file_id].annPredictionParser.getPredictions(ctx)) {
                        tt.addAll(pred.getSOTerms());
                    }
                    return tt;
                }
            });
        }
        for (String s : commonSamples) {
            venn1List.add(new VennGType(s));
        }
        /* START : digest results ====================== */
        Counter<String> diff = new Counter<String>();
        List<LineAndFile> row = new ArrayList<LineAndFile>();
        CloseableIterator<LineAndFile> iter = variants.iterator();
        for (; ; ) {
            LineAndFile rec = null;
            if (iter.hasNext()) {
                rec = iter.next();
            }
            if (rec == null || (!row.isEmpty() && varcmp.compare(row.get(0), rec) != 0)) {
                if (!row.isEmpty()) {
                    diff.incr("count.variations");
                    VariantContext[] contexes_init = new VariantContext[] { null, null };
                    for (LineAndFile var : row) {
                        if (contexes_init[var.fileIdx] != null) {
                            LOG.error("Duplicate context in " + inputs[var.fileIdx].filename + " : " + var.line);
                            continue;
                        }
                        contexes_init[var.fileIdx] = var.getContext();
                    }
                    for (Venn0 venn : venn1List) {
                        venn.visit(contexes_init);
                    }
                    row.clear();
                }
                if (rec == null)
                    break;
            }
            row.add(rec);
        }
        iter.close();
        /* END : digest results ====================== */
        pw = super.openFileOrStdoutAsPrintWriter(outputFile);
        XMLOutputFactory xmlfactory = XMLOutputFactory.newInstance();
        w = xmlfactory.createXMLStreamWriter(pw);
        w.writeStartElement("html");
        w.writeStartElement("body");
        /* specific samples */
        w.writeStartElement("div");
        w.writeStartElement("dl");
        for (int i = 0; i < 3; ++i) {
            String title;
            Set<String> samples;
            switch(i) {
                case 0:
                case 1:
                    title = "Sample(s) for " + this.inputs[i].filename + ".";
                    samples = new TreeSet<String>(this.inputs[i].header.getSampleNamesInOrder());
                    samples.removeAll(commonSamples);
                    break;
                default:
                    title = "Common Sample(s).";
                    samples = new TreeSet<String>(commonSamples);
                    break;
            }
            w.writeStartElement("dt");
            w.writeCharacters(title);
            w.writeEndElement();
            w.writeStartElement("dd");
            w.writeStartElement("ol");
            for (String s : samples) {
                w.writeStartElement("li");
                w.writeCharacters(s);
                w.writeEndElement();
            }
            w.writeEndElement();
            w.writeEndElement();
        }
        // dl
        w.writeEndElement();
        // div
        w.writeEndElement();
        for (Venn0 v : venn1List) {
            v.write(w);
        }
        // body
        w.writeEndElement();
        // html
        w.writeEndElement();
        w.writeEndDocument();
        w.close();
        w = null;
        pw.flush();
        pw.close();
        pw = null;
    } catch (Exception err) {
        LOG.error(err);
        return -1;
    } finally {
        CloserUtil.close(w);
        CloserUtil.close(pw);
        if (variants != null)
            variants.cleanup();
    }
    return 0;
}

Example 4 with SnpEffPredictionParserFactory

use of com.github.lindenb.jvarkit.util.vcf.predictions.SnpEffPredictionParserFactory in project jvarkit by lindenb.

the class VcfGeneOntology method filterVcfIterator.

private void filterVcfIterator(final VcfIterator in) throws IOException {
    VariantContextWriter w = null;
    try {
        VCFHeader header = in.getHeader();
        VCFHeader h2 = new VCFHeader(header);
        h2.addMetaDataLine(new VCFInfoHeaderLine(TAG, VCFHeaderLineCount.UNBOUNDED, VCFHeaderLineType.String, "GO terms from GO " + GO + " and GOA=" + GOA));
        h2.addMetaDataLine(new VCFHeaderLine(getClass().getSimpleName() + "CmdLine", String.valueOf(getProgramCommandLine())));
        h2.addMetaDataLine(new VCFHeaderLine(getClass().getSimpleName() + "Version", String.valueOf(getVersion())));
        h2.addMetaDataLine(new VCFHeaderLine(getClass().getSimpleName() + "HtsJdkVersion", HtsjdkVersion.getVersion()));
        h2.addMetaDataLine(new VCFHeaderLine(getClass().getSimpleName() + "HtsJdkHome", HtsjdkVersion.getHome()));
        if (filterName != null) {
            h2.addMetaDataLine(new VCFFilterHeaderLine(filterName, "Flag  GO terms " + (inverse_filter ? " not descendant of " : "") + " the provided GO terms"));
        }
        w = super.openVariantContextWriter(outputFile);
        w.writeHeader(h2);
        final SAMSequenceDictionaryProgress progess = new SAMSequenceDictionaryProgress(header.getSequenceDictionary());
        final SnpEffPredictionParser snpEffPredictionParser = new SnpEffPredictionParserFactory().header(header).get();
        final VepPredictionParser vepPredictionParser = new VepPredictionParserFactory().header(header).get();
        while (in.hasNext()) {
            if (System.out.checkError())
                break;
            VariantContext ctx = progess.watch(in.next());
            /* symbols for this variant */
            Set<String> symbols = new HashSet<String>();
            /* scan SNPEFF gene */
            for (SnpEffPrediction pred : snpEffPredictionParser.getPredictions(ctx)) {
                String genName = pred.getGeneName();
                if (genName == null || genName.isEmpty())
                    continue;
                symbols.add(genName);
            }
            /* scan VEP gene */
            for (VepPrediction pred : vepPredictionParser.getPredictions(ctx)) {
                String genName = pred.getGeneName();
                if (!(genName == null || genName.isEmpty())) {
                    symbols.add(genName);
                }
                genName = pred.getGene();
                if (!(genName == null || genName.isEmpty())) {
                    symbols.add(genName);
                }
                genName = pred.getHGNC();
                if (!(genName == null || genName.isEmpty())) {
                    symbols.add(genName);
                }
            }
            /* only keep known GENES from GOA */
            symbols.retainAll(this.name2go.keySet());
            boolean found_child_of_filter = false;
            /* ATTS */
            List<String> atts = new ArrayList<String>();
            /* loop over symbols */
            for (String symbol : symbols) {
                /* go terms associated to this symbol */
                Set<GoTree.Term> t2 = this.name2go.get(symbol);
                if (t2 == null || t2.isEmpty())
                    continue;
                StringBuilder sb = new StringBuilder(symbol);
                sb.append("|");
                boolean first = true;
                for (GoTree.Term gt : t2) {
                    /* user gave terms to filter */
                    if (!found_child_of_filter && this.goTermToFilter != null) {
                        for (GoTree.Term userTerm : this.goTermToFilter) {
                            if (userTerm.hasDescendant(gt.getAcn())) {
                                found_child_of_filter = true;
                                break;
                            }
                        }
                    }
                    if (!first)
                        sb.append("&");
                    sb.append(gt.getAcn());
                    first = false;
                }
                atts.add(sb.toString());
            }
            /* no go term was found */
            if (atts.isEmpty()) {
                if (!removeIfNoGo) {
                    w.add(ctx);
                }
                continue;
            }
            VariantContextBuilder vcb = new VariantContextBuilder(ctx);
            /* check children of user's terms */
            if (this.goTermToFilter != null) {
                /* keep if found children*/
                if ((this.inverse_filter && found_child_of_filter) || (!this.inverse_filter && !found_child_of_filter)) {
                    /* don't remove, but set filter */
                    if (this.filterName != null) {
                        Set<String> filters = new HashSet<String>(ctx.getFilters());
                        filters.add(this.filterName);
                        vcb.filters(filters);
                    } else {
                        continue;
                    }
                }
            }
            /* add go terms */
            vcb.attribute(this.TAG, atts);
            w.add(vcb.make());
        }
        progess.finish();
        w.close();
        w = null;
    } finally {
        CloserUtil.close(w);
        w = null;
    }
}