Clover icon

jalviewX

  1. Project Clover database Wed Oct 31 2018 15:13:58 GMT
  2. Package jalview.io.vcf

File VCFLoader.java

 

Coverage histogram

../../../img/srcFileCovDistChart7.png
28% of files have more coverage

Code metrics

156
358
33
2
1,474
807
136
0.38
10.85
16.5
4.12

Classes

Class Line # Actions
VCFLoader 52 355 134 176
0.6752767667.5%
VCFLoader.VCFMap 64 3 2 2
0.660%
 

Contributing tests

No tests hitting this source file were found.

Source view

1    package jalview.io.vcf;
2   
3    import jalview.analysis.AlignmentUtils;
4    import jalview.analysis.Dna;
5    import jalview.api.AlignViewControllerGuiI;
6    import jalview.bin.Cache;
7    import jalview.datamodel.DBRefEntry;
8    import jalview.datamodel.GeneLociI;
9    import jalview.datamodel.Mapping;
10    import jalview.datamodel.SequenceFeature;
11    import jalview.datamodel.SequenceI;
12    import jalview.datamodel.features.FeatureAttributeType;
13    import jalview.datamodel.features.FeatureSource;
14    import jalview.datamodel.features.FeatureSources;
15    import jalview.ext.ensembl.EnsemblMap;
16    import jalview.ext.htsjdk.HtsContigDb;
17    import jalview.ext.htsjdk.VCFReader;
18    import jalview.io.gff.Gff3Helper;
19    import jalview.io.gff.SequenceOntologyI;
20    import jalview.util.MapList;
21    import jalview.util.MappingUtils;
22    import jalview.util.MessageManager;
23   
24    import java.io.File;
25    import java.io.IOException;
26    import java.util.ArrayList;
27    import java.util.HashMap;
28    import java.util.List;
29    import java.util.Map;
30    import java.util.Map.Entry;
31    import java.util.regex.Pattern;
32    import java.util.regex.PatternSyntaxException;
33   
34    import htsjdk.samtools.SAMException;
35    import htsjdk.samtools.SAMSequenceDictionary;
36    import htsjdk.samtools.SAMSequenceRecord;
37    import htsjdk.samtools.util.CloseableIterator;
38    import htsjdk.variant.variantcontext.Allele;
39    import htsjdk.variant.variantcontext.VariantContext;
40    import htsjdk.variant.vcf.VCFHeader;
41    import htsjdk.variant.vcf.VCFHeaderLine;
42    import htsjdk.variant.vcf.VCFHeaderLineCount;
43    import htsjdk.variant.vcf.VCFHeaderLineType;
44    import htsjdk.variant.vcf.VCFInfoHeaderLine;
45   
46    /**
47    * A class to read VCF data (using the htsjdk) and add variants as sequence
48    * features on dna and any related protein product sequences
49    *
50    * @author gmcarstairs
51    */
 
52    public class VCFLoader
53    {
54    /**
55    * A class to model the mapping from sequence to VCF coordinates. Cases include
56    * <ul>
57    * <li>a direct 1:1 mapping where the sequence is one of the VCF contigs</li>
58    * <li>a mapping of sequence to chromosomal coordinates, where sequence and VCF
59    * use the same reference assembly</li>
60    * <li>a modified mapping of sequence to chromosomal coordinates, where sequence
61    * and VCF use different reference assembles</li>
62    * </ul>
63    */
 
64    class VCFMap
65    {
66    final String chromosome;
67   
68    final MapList map;
69   
 
70  24 toggle VCFMap(String chr, MapList m)
71    {
72  24 chromosome = chr;
73  24 map = m;
74    }
75   
 
76  0 toggle @Override
77    public String toString()
78    {
79  0 return chromosome + ":" + map.toString();
80    }
81    }
82   
83    /*
84    * Lookup keys, and default values, for Preference entries that describe
85    * patterns for VCF and VEP fields to capture
86    */
87    private static final String VEP_FIELDS_PREF = "VEP_FIELDS";
88   
89    private static final String VCF_FIELDS_PREF = "VCF_FIELDS";
90   
91    private static final String DEFAULT_VCF_FIELDS = ".*";
92   
93    private static final String DEFAULT_VEP_FIELDS = ".*";// "Allele,Consequence,IMPACT,SWISSPROT,SIFT,PolyPhen,CLIN_SIG";
94   
95    /*
96    * keys to fields of VEP CSQ consequence data
97    * see https://www.ensembl.org/info/docs/tools/vep/vep_formats.html
98    */
99    private static final String CSQ_CONSEQUENCE_KEY = "Consequence";
100    private static final String CSQ_ALLELE_KEY = "Allele";
101    private static final String CSQ_ALLELE_NUM_KEY = "ALLELE_NUM"; // 0 (ref), 1...
102    private static final String CSQ_FEATURE_KEY = "Feature"; // Ensembl stable id
103   
104    /*
105    * default VCF INFO key for VEP consequence data
106    * NB this can be overridden running VEP with --vcf_info_field
107    * - we don't handle this case (require identifier to be CSQ)
108    */
109    private static final String CSQ_FIELD = "CSQ";
110   
111    /*
112    * separator for fields in consequence data is '|'
113    */
114    private static final String PIPE_REGEX = "\\|";
115   
116    /*
117    * key for Allele Frequency output by VEP
118    * see http://www.ensembl.org/info/docs/tools/vep/vep_formats.html
119    */
120    private static final String ALLELE_FREQUENCY_KEY = "AF";
121   
122    /*
123    * delimiter that separates multiple consequence data blocks
124    */
125    private static final String COMMA = ",";
126   
127    /*
128    * the feature group assigned to a VCF variant in Jalview
129    */
130    private static final String FEATURE_GROUP_VCF = "VCF";
131   
132    /*
133    * internal delimiter used to build keys for assemblyMappings
134    *
135    */
136    private static final String EXCL = "!";
137   
138    /*
139    * the VCF file we are processing
140    */
141    protected String vcfFilePath;
142   
143    /*
144    * mappings between VCF and sequence reference assembly regions, as
145    * key = "species!chromosome!fromAssembly!toAssembly
146    * value = Map{fromRange, toRange}
147    */
148    private Map<String, Map<int[], int[]>> assemblyMappings;
149   
150    private VCFReader reader;
151   
152    /*
153    * holds details of the VCF header lines (metadata)
154    */
155    private VCFHeader header;
156   
157    /*
158    * a Dictionary of contigs (if present) referenced in the VCF file
159    */
160    private SAMSequenceDictionary dictionary;
161   
162    /*
163    * the position (0...) of field in each block of
164    * CSQ (consequence) data (if declared in the VCF INFO header for CSQ)
165    * see http://www.ensembl.org/info/docs/tools/vep/vep_formats.html
166    */
167    private int csqConsequenceFieldIndex = -1;
168    private int csqAlleleFieldIndex = -1;
169    private int csqAlleleNumberFieldIndex = -1;
170    private int csqFeatureFieldIndex = -1;
171   
172    // todo the same fields for SnpEff ANN data if wanted
173    // see http://snpeff.sourceforge.net/SnpEff_manual.html#input
174   
175    /*
176    * a unique identifier under which to save metadata about feature
177    * attributes (selected INFO field data)
178    */
179    private String sourceId;
180   
181    /*
182    * The INFO IDs of data that is both present in the VCF file, and
183    * also matched by any filters for data of interest
184    */
185    List<String> vcfFieldsOfInterest;
186   
187    /*
188    * The field offsets and identifiers for VEP (CSQ) data that is both present
189    * in the VCF file, and also matched by any filters for data of interest
190    * for example 0 -> Allele, 1 -> Consequence, ..., 36 -> SIFT, ...
191    */
192    Map<Integer, String> vepFieldsOfInterest;
193   
194    /**
195    * Constructor given a VCF file
196    *
197    * @param alignment
198    */
 
199  4 toggle public VCFLoader(String vcfFile)
200    {
201  4 try
202    {
203  4 initialise(vcfFile);
204    } catch (IOException e)
205    {
206  0 System.err.println("Error opening VCF file: " + e.getMessage());
207    }
208   
209    // map of species!chromosome!fromAssembly!toAssembly to {fromRange, toRange}
210  4 assemblyMappings = new HashMap<>();
211    }
212   
213    /**
214    * Starts a new thread to query and load VCF variant data on to the given
215    * sequences
216    * <p>
217    * This method is not thread safe - concurrent threads should use separate
218    * instances of this class.
219    *
220    * @param seqs
221    * @param gui
222    */
 
223  0 toggle public void loadVCF(SequenceI[] seqs, final AlignViewControllerGuiI gui)
224    {
225  0 if (gui != null)
226    {
227  0 gui.setStatus(MessageManager.getString("label.searching_vcf"));
228    }
229   
230  0 new Thread()
231    {
 
232  0 toggle @Override
233    public void run()
234    {
235  0 VCFLoader.this.doLoad(seqs, gui);
236    }
237    }.start();
238    }
239   
240    /**
241    * Reads the specified contig sequence and adds its VCF variants to it
242    *
243    * @param contig
244    * the id of a single sequence (contig) to load
245    * @return
246    */
 
247  3 toggle public SequenceI loadVCFContig(String contig)
248    {
249  3 String ref = header.getOtherHeaderLine(VCFHeader.REFERENCE_KEY)
250    .getValue();
251  3 if (ref.startsWith("file://"))
252    {
253  0 ref = ref.substring(7);
254    }
255   
256  3 SequenceI seq = null;
257  3 File dbFile = new File(ref);
258   
259  3 if (dbFile.exists())
260    {
261  3 HtsContigDb db = new HtsContigDb("", dbFile);
262  3 seq = db.getSequenceProxy(contig);
263  3 loadSequenceVCF(seq, ref);
264  3 db.close();
265    }
266    else
267    {
268  0 System.err.println("VCF reference not found: " + ref);
269    }
270   
271  3 return seq;
272    }
273   
274    /**
275    * Loads VCF on to one or more sequences
276    *
277    * @param seqs
278    * @param gui
279    * optional callback handler for messages
280    */
 
281  3 toggle protected void doLoad(SequenceI[] seqs, AlignViewControllerGuiI gui)
282    {
283  3 try
284    {
285  3 VCFHeaderLine ref = header
286    .getOtherHeaderLine(VCFHeader.REFERENCE_KEY);
287  3 String vcfAssembly = ref.getValue();
288   
289  3 int varCount = 0;
290  3 int seqCount = 0;
291   
292    /*
293    * query for VCF overlapping each sequence in turn
294    */
295  3 for (SequenceI seq : seqs)
296    {
297  21 int added = loadSequenceVCF(seq, vcfAssembly);
298  21 if (added > 0)
299    {
300  11 seqCount++;
301  11 varCount += added;
302  11 transferAddedFeatures(seq);
303    }
304    }
305  3 if (gui != null)
306    {
307  0 String msg = MessageManager.formatMessage("label.added_vcf",
308    varCount, seqCount);
309  0 gui.setStatus(msg);
310  0 if (gui.getFeatureSettingsUI() != null)
311    {
312  0 gui.getFeatureSettingsUI().discoverAllFeatureData();
313    }
314    }
315    } catch (Throwable e)
316    {
317  0 System.err.println("Error processing VCF: " + e.getMessage());
318  0 e.printStackTrace();
319  0 if (gui != null)
320    {
321  0 gui.setStatus("Error occurred - see console for details");
322    }
323    } finally
324    {
325  3 if (reader != null)
326    {
327  3 try
328    {
329  3 reader.close();
330    } catch (IOException e)
331    {
332    // ignore
333    }
334    }
335  3 header = null;
336  3 dictionary = null;
337    }
338    }
339   
340    /**
341    * Opens the VCF file and parses header data
342    *
343    * @param filePath
344    * @throws IOException
345    */
 
346  4 toggle private void initialise(String filePath) throws IOException
347    {
348  4 vcfFilePath = filePath;
349   
350  4 reader = new VCFReader(filePath);
351   
352  4 header = reader.getFileHeader();
353   
354  4 try
355    {
356  4 dictionary = header.getSequenceDictionary();
357    } catch (SAMException e)
358    {
359    // ignore - thrown if any contig line lacks length info
360    }
361   
362  4 sourceId = filePath;
363   
364  4 saveMetadata(sourceId);
365   
366    /*
367    * get offset of CSQ ALLELE_NUM and Feature if declared
368    */
369  4 parseCsqHeader();
370    }
371   
372    /**
373    * Reads metadata (such as INFO field descriptions and datatypes) and saves
374    * them for future reference
375    *
376    * @param theSourceId
377    */
 
378  4 toggle void saveMetadata(String theSourceId)
379    {
380  4 List<Pattern> vcfFieldPatterns = getFieldMatchers(VCF_FIELDS_PREF,
381    DEFAULT_VCF_FIELDS);
382  4 vcfFieldsOfInterest = new ArrayList<>();
383   
384  4 FeatureSource metadata = new FeatureSource(theSourceId);
385   
386  4 for (VCFInfoHeaderLine info : header.getInfoHeaderLines())
387    {
388  9 String attributeId = info.getID();
389  9 String desc = info.getDescription();
390  9 VCFHeaderLineType type = info.getType();
391  9 FeatureAttributeType attType = null;
392  9 switch (type)
393    {
394  0 case Character:
395  0 attType = FeatureAttributeType.Character;
396  0 break;
397  0 case Flag:
398  0 attType = FeatureAttributeType.Flag;
399  0 break;
400  5 case Float:
401  5 attType = FeatureAttributeType.Float;
402  5 break;
403  3 case Integer:
404  3 attType = FeatureAttributeType.Integer;
405  3 break;
406  1 case String:
407  1 attType = FeatureAttributeType.String;
408  1 break;
409    }
410  9 metadata.setAttributeName(attributeId, desc);
411  9 metadata.setAttributeType(attributeId, attType);
412   
413  9 if (isFieldWanted(attributeId, vcfFieldPatterns))
414    {
415  9 vcfFieldsOfInterest.add(attributeId);
416    }
417    }
418   
419  4 FeatureSources.getInstance().addSource(theSourceId, metadata);
420    }
421   
422    /**
423    * Answers true if the field id is matched by any of the filter patterns, else
424    * false. Matching is against regular expression patterns, and is not
425    * case-sensitive.
426    *
427    * @param id
428    * @param filters
429    * @return
430    */
 
431  18 toggle private boolean isFieldWanted(String id, List<Pattern> filters)
432    {
433  18 for (Pattern p : filters)
434    {
435  18 if (p.matcher(id.toUpperCase()).matches())
436    {
437  18 return true;
438    }
439    }
440  0 return false;
441    }
442   
443    /**
444    * Records 'wanted' fields defined in the CSQ INFO header (if there is one).
445    * Also records the position of selected fields (Allele, ALLELE_NUM, Feature)
446    * required for processing.
447    * <p>
448    * CSQ fields are declared in the CSQ INFO Description e.g.
449    * <p>
450    * Description="Consequence ...from ... VEP. Format: Allele|Consequence|...
451    */
 
452  4 toggle protected void parseCsqHeader()
453    {
454  4 List<Pattern> vepFieldFilters = getFieldMatchers(VEP_FIELDS_PREF,
455    DEFAULT_VEP_FIELDS);
456  4 vepFieldsOfInterest = new HashMap<>();
457   
458  4 VCFInfoHeaderLine csqInfo = header.getInfoHeaderLine(CSQ_FIELD);
459  4 if (csqInfo == null)
460    {
461  3 return;
462    }
463   
464    /*
465    * parse out the pipe-separated list of CSQ fields; we assume here that
466    * these form the last part of the description, and contain no spaces
467    */
468  1 String desc = csqInfo.getDescription();
469  1 int spacePos = desc.lastIndexOf(" ");
470  1 desc = desc.substring(spacePos + 1);
471   
472  1 if (desc != null)
473    {
474  1 String[] format = desc.split(PIPE_REGEX);
475  1 int index = 0;
476  1 for (String field : format)
477    {
478  9 if (CSQ_CONSEQUENCE_KEY.equals(field))
479    {
480  1 csqConsequenceFieldIndex = index;
481    }
482  9 if (CSQ_ALLELE_NUM_KEY.equals(field))
483    {
484  0 csqAlleleNumberFieldIndex = index;
485    }
486  9 if (CSQ_ALLELE_KEY.equals(field))
487    {
488  1 csqAlleleFieldIndex = index;
489    }
490  9 if (CSQ_FEATURE_KEY.equals(field))
491    {
492  1 csqFeatureFieldIndex = index;
493    }
494   
495  9 if (isFieldWanted(field, vepFieldFilters))
496    {
497  9 vepFieldsOfInterest.put(index, field);
498    }
499   
500  9 index++;
501    }
502    }
503    }
504   
505    /**
506    * Reads the Preference value for the given key, with default specified if no
507    * preference set. The value is interpreted as a comma-separated list of
508    * regular expressions, and converted into a list of compiled patterns ready
509    * for matching. Patterns are forced to upper-case for non-case-sensitive
510    * matching.
511    * <p>
512    * This supports user-defined filters for fields of interest to capture while
513    * processing data. For example, VCF_FIELDS = AF,AC* would mean that VCF INFO
514    * fields with an ID of AF, or starting with AC, would be matched.
515    *
516    * @param key
517    * @param def
518    * @return
519    */
 
520  8 toggle private List<Pattern> getFieldMatchers(String key, String def)
521    {
522  8 String pref = Cache.getDefault(key, def);
523  8 List<Pattern> patterns = new ArrayList<>();
524  8 String[] tokens = pref.split(",");
525  8 for (String token : tokens)
526    {
527  8 try
528    {
529  8 patterns.add(Pattern.compile(token.toUpperCase()));
530    } catch (PatternSyntaxException e)
531    {
532  0 System.err.println("Invalid pattern ignored: " + token);
533    }
534    }
535  8 return patterns;
536    }
537   
538    /**
539    * Transfers VCF features to sequences to which this sequence has a mapping.
540    * If the mapping is 3:1, computes peptide variants from nucleotide variants.
541    *
542    * @param seq
543    */
 
544  11 toggle protected void transferAddedFeatures(SequenceI seq)
545    {
546  11 DBRefEntry[] dbrefs = seq.getDBRefs();
547  11 if (dbrefs == null)
548    {
549  0 return;
550    }
551  11 for (DBRefEntry dbref : dbrefs)
552    {
553  16 Mapping mapping = dbref.getMap();
554  16 if (mapping == null || mapping.getTo() == null)
555    {
556  11 continue;
557    }
558   
559  5 SequenceI mapTo = mapping.getTo();
560  5 MapList map = mapping.getMap();
561  5 if (map.getFromRatio() == 3)
562    {
563    /*
564    * dna-to-peptide product mapping
565    */
566  5 AlignmentUtils.computeProteinFeatures(seq, mapTo, map);
567    }
568    else
569    {
570    /*
571    * nucleotide-to-nucleotide mapping e.g. transcript to CDS
572    */
573  0 List<SequenceFeature> features = seq.getFeatures()
574    .getPositionalFeatures(SequenceOntologyI.SEQUENCE_VARIANT);
575  0 for (SequenceFeature sf : features)
576    {
577  0 if (FEATURE_GROUP_VCF.equals(sf.getFeatureGroup()))
578    {
579  0 transferFeature(sf, mapTo, map);
580    }
581    }
582    }
583    }
584    }
585   
586    /**
587    * Tries to add overlapping variants read from a VCF file to the given sequence,
588    * and returns the number of variant features added
589    *
590    * @param seq
591    * @param vcfAssembly
592    * @return
593    */
 
594  24 toggle protected int loadSequenceVCF(SequenceI seq, String vcfAssembly)
595    {
596  24 VCFMap vcfMap = getVcfMap(seq, vcfAssembly);
597  24 if (vcfMap == null)
598    {
599  0 return 0;
600    }
601   
602    /*
603    * work with the dataset sequence here
604    */
605  24 SequenceI dss = seq.getDatasetSequence();
606  24 if (dss == null)
607    {
608  3 dss = seq;
609    }
610  24 return addVcfVariants(dss, vcfMap);
611    }
612   
613    /**
614    * Answers a map from sequence coordinates to VCF chromosome ranges
615    *
616    * @param seq
617    * @param vcfAssembly
618    * @return
619    */
 
620  24 toggle private VCFMap getVcfMap(SequenceI seq, String vcfAssembly)
621    {
622    /*
623    * simplest case: sequence has id and length matching a VCF contig
624    */
625  24 VCFMap vcfMap = null;
626  24 if (dictionary != null)
627    {
628  3 vcfMap = getContigMap(seq);
629    }
630  24 if (vcfMap != null)
631    {
632  3 return vcfMap;
633    }
634   
635    /*
636    * otherwise, map to VCF from chromosomal coordinates
637    * of the sequence (if known)
638    */
639  21 GeneLociI seqCoords = seq.getGeneLoci();
640  21 if (seqCoords == null)
641    {
642  0 Cache.log.warn(String.format(
643    "Can't query VCF for %s as chromosome coordinates not known",
644    seq.getName()));
645  0 return null;
646    }
647   
648  21 String species = seqCoords.getSpeciesId();
649  21 String chromosome = seqCoords.getChromosomeId();
650  21 String seqRef = seqCoords.getAssemblyId();
651  21 MapList map = seqCoords.getMap();
652   
653  21 if (!vcfSpeciesMatchesSequence(vcfAssembly, species))
654    {
655  0 return null;
656    }
657   
658  21 if (vcfAssemblyMatchesSequence(vcfAssembly, seqRef))
659    {
660  21 return new VCFMap(chromosome, map);
661    }
662   
663  0 if (!"GRCh38".equalsIgnoreCase(seqRef) // Ensembl
664    || !vcfAssembly.contains("Homo_sapiens_assembly19")) // gnomAD
665    {
666  0 return null;
667    }
668   
669    /*
670    * map chromosomal coordinates from sequence to VCF if the VCF
671    * data has a different reference assembly to the sequence
672    */
673    // TODO generalise for cases other than GRCh38 -> GRCh37 !
674    // - or get the user to choose in a dialog
675   
676  0 List<int[]> toVcfRanges = new ArrayList<>();
677  0 List<int[]> fromSequenceRanges = new ArrayList<>();
678  0 String toRef = "GRCh37";
679   
680  0 for (int[] range : map.getToRanges())
681    {
682  0 int[] fromRange = map.locateInFrom(range[0], range[1]);
683  0 if (fromRange == null)
684    {
685    // corrupted map?!?
686  0 continue;
687    }
688   
689  0 int[] newRange = mapReferenceRange(range, chromosome, "human", seqRef,
690    toRef);
691  0 if (newRange == null)
692    {
693  0 Cache.log.error(
694    String.format("Failed to map %s:%s:%s:%d:%d to %s", species,
695    chromosome, seqRef, range[0], range[1], toRef));
696  0 continue;
697    }
698    else
699    {
700  0 toVcfRanges.add(newRange);
701  0 fromSequenceRanges.add(fromRange);
702    }
703    }
704   
705  0 return new VCFMap(chromosome,
706    new MapList(fromSequenceRanges, toVcfRanges, 1, 1));
707    }
708   
709    /**
710    * If the sequence id matches a contig declared in the VCF file, and the
711    * sequence length matches the contig length, then returns a 1:1 map of the
712    * sequence to the contig, else returns null
713    *
714    * @param seq
715    * @return
716    */
 
717  3 toggle private VCFMap getContigMap(SequenceI seq)
718    {
719  3 String id = seq.getName();
720  3 SAMSequenceRecord contig = dictionary.getSequence(id);
721  3 if (contig != null)
722    {
723  3 int len = seq.getLength();
724  3 if (len == contig.getSequenceLength())
725    {
726  3 MapList map = new MapList(new int[] { 1, len },
727    new int[]
728    { 1, len }, 1, 1);
729  3 return new VCFMap(id, map);
730    }
731    }
732  0 return null;
733    }
734   
735    /**
736    * Answers true if we determine that the VCF data uses the same reference
737    * assembly as the sequence, else false
738    *
739    * @param vcfAssembly
740    * @param seqRef
741    * @return
742    */
 
743  21 toggle private boolean vcfAssemblyMatchesSequence(String vcfAssembly,
744    String seqRef)
745    {
746    // TODO improve on this stub, which handles gnomAD and
747    // hopes for the best for other cases
748   
749  21 if ("GRCh38".equalsIgnoreCase(seqRef) // Ensembl
750    && vcfAssembly.contains("Homo_sapiens_assembly19")) // gnomAD
751    {
752  0 return false;
753    }
754  21 return true;
755    }
756   
757    /**
758    * Answers true if the species inferred from the VCF reference identifier
759    * matches that for the sequence
760    *
761    * @param vcfAssembly
762    * @param speciesId
763    * @return
764    */
 
765  21 toggle boolean vcfSpeciesMatchesSequence(String vcfAssembly, String speciesId)
766    {
767    // PROBLEM 1
768    // there are many aliases for species - how to equate one with another?
769    // PROBLEM 2
770    // VCF ##reference header is an unstructured URI - how to extract species?
771    // perhaps check if ref includes any (Ensembl) alias of speciesId??
772    // TODO ask the user to confirm this??
773   
774  21 if (vcfAssembly.contains("Homo_sapiens") // gnomAD exome data example
775    && "HOMO_SAPIENS".equals(speciesId)) // Ensembl species id
776    {
777  21 return true;
778    }
779   
780  0 if (vcfAssembly.contains("c_elegans") // VEP VCF response example
781    && "CAENORHABDITIS_ELEGANS".equals(speciesId)) // Ensembl
782    {
783  0 return true;
784    }
785   
786    // this is not a sustainable solution...
787   
788  0 return false;
789    }
790   
791    /**
792    * Queries the VCF reader for any variants that overlap the mapped chromosome
793    * ranges of the sequence, and adds as variant features. Returns the number of
794    * overlapping variants found.
795    *
796    * @param seq
797    * @param map
798    * mapping from sequence to VCF coordinates
799    * @return
800    */
 
801  24 toggle protected int addVcfVariants(SequenceI seq, VCFMap map)
802    {
803  24 boolean forwardStrand = map.map.isToForwardStrand();
804   
805    /*
806    * query the VCF for overlaps of each contiguous chromosomal region
807    */
808  24 int count = 0;
809   
810  24 for (int[] range : map.map.getToRanges())
811    {
812  39 int vcfStart = Math.min(range[0], range[1]);
813  39 int vcfEnd = Math.max(range[0], range[1]);
814  39 CloseableIterator<VariantContext> variants = reader
815    .query(map.chromosome, vcfStart, vcfEnd);
816  67 while (variants.hasNext())
817    {
818  28 VariantContext variant = variants.next();
819   
820  28 int[] featureRange = map.map.locateInFrom(variant.getStart(),
821    variant.getEnd());
822   
823  28 if (featureRange != null)
824    {
825  26 int featureStart = Math.min(featureRange[0], featureRange[1]);
826  26 int featureEnd = Math.max(featureRange[0], featureRange[1]);
827  26 count += addAlleleFeatures(seq, variant, featureStart, featureEnd,
828    forwardStrand);
829    }
830    }
831  39 variants.close();
832    }
833   
834  24 return count;
835    }
836   
837    /**
838    * A convenience method to get the AF value for the given alternate allele
839    * index
840    *
841    * @param variant
842    * @param alleleIndex
843    * @return
844    */
 
845  43 toggle protected float getAlleleFrequency(VariantContext variant, int alleleIndex)
846    {
847  43 float score = 0f;
848  43 String attributeValue = getAttributeValue(variant,
849    ALLELE_FREQUENCY_KEY, alleleIndex);
850  43 if (attributeValue != null)
851    {
852  43 try
853    {
854  43 score = Float.parseFloat(attributeValue);
855    } catch (NumberFormatException e)
856    {
857    // leave as 0
858    }
859    }
860   
861  43 return score;
862    }
863   
864    /**
865    * A convenience method to get an attribute value for an alternate allele
866    *
867    * @param variant
868    * @param attributeName
869    * @param alleleIndex
870    * @return
871    */
 
872  119 toggle protected String getAttributeValue(VariantContext variant,
873    String attributeName, int alleleIndex)
874    {
875  119 Object att = variant.getAttribute(attributeName);
876   
877  119 if (att instanceof String)
878    {
879  40 return (String) att;
880    }
881  79 else if (att instanceof ArrayList)
882    {
883  79 return ((List<String>) att).get(alleleIndex);
884    }
885   
886  0 return null;
887    }
888   
889    /**
890    * Adds one variant feature for each allele in the VCF variant record, and
891    * returns the number of features added.
892    *
893    * @param seq
894    * @param variant
895    * @param featureStart
896    * @param featureEnd
897    * @param forwardStrand
898    * @return
899    */
 
900  26 toggle protected int addAlleleFeatures(SequenceI seq, VariantContext variant,
901    int featureStart, int featureEnd, boolean forwardStrand)
902    {
903  26 int added = 0;
904   
905    /*
906    * Javadoc says getAlternateAlleles() imposes no order on the list returned
907    * so we proceed defensively to get them in strict order
908    */
909  26 int altAlleleCount = variant.getAlternateAlleles().size();
910  69 for (int i = 0; i < altAlleleCount; i++)
911    {
912  43 added += addAlleleFeature(seq, variant, i, featureStart, featureEnd,
913    forwardStrand);
914    }
915  26 return added;
916    }
917   
918    /**
919    * Inspects one allele and attempts to add a variant feature for it to the
920    * sequence. The additional data associated with this allele is extracted to
921    * store in the feature's key-value map. Answers the number of features added (0
922    * or 1).
923    *
924    * @param seq
925    * @param variant
926    * @param altAlleleIndex
927    * (0, 1..)
928    * @param featureStart
929    * @param featureEnd
930    * @param forwardStrand
931    * @return
932    */
 
933  43 toggle protected int addAlleleFeature(SequenceI seq, VariantContext variant,
934    int altAlleleIndex, int featureStart, int featureEnd,
935    boolean forwardStrand)
936    {
937  43 String reference = variant.getReference().getBaseString();
938  43 Allele alt = variant.getAlternateAllele(altAlleleIndex);
939  43 String allele = alt.getBaseString();
940   
941    /*
942    * insertion after a genomic base, if on reverse strand, has to be
943    * converted to insertion of complement after the preceding position
944    */
945  43 int referenceLength = reference.length();
946  43 if (!forwardStrand && allele.length() > referenceLength
947    && allele.startsWith(reference))
948    {
949  4 featureStart -= referenceLength;
950  4 featureEnd = featureStart;
951  4 char insertAfter = seq.getCharAt(featureStart - seq.getStart());
952  4 reference = Dna.reverseComplement(String.valueOf(insertAfter));
953  4 allele = allele.substring(referenceLength) + reference;
954    }
955   
956    /*
957    * build the ref,alt allele description e.g. "G,A", using the base
958    * complement if the sequence is on the reverse strand
959    */
960  43 StringBuilder sb = new StringBuilder();
961  43 sb.append(forwardStrand ? reference : Dna.reverseComplement(reference));
962  43 sb.append(COMMA);
963  43 sb.append(forwardStrand ? allele : Dna.reverseComplement(allele));
964  43 String alleles = sb.toString(); // e.g. G,A
965   
966    /*
967    * pick out the consequence data (if any) that is for the current allele
968    * and feature (transcript) that matches the current sequence
969    */
970  43 String consequence = getConsequenceForAlleleAndFeature(variant, CSQ_FIELD,
971    altAlleleIndex, csqAlleleFieldIndex,
972    csqAlleleNumberFieldIndex, seq.getName().toLowerCase(),
973    csqFeatureFieldIndex);
974   
975    /*
976    * pick out the ontology term for the consequence type
977    */
978  43 String type = SequenceOntologyI.SEQUENCE_VARIANT;
979  43 if (consequence != null)
980    {
981  7 type = getOntologyTerm(consequence);
982    }
983   
984  43 float score = getAlleleFrequency(variant, altAlleleIndex);
985   
986  43 SequenceFeature sf = new SequenceFeature(type, alleles, featureStart,
987    featureEnd, score, FEATURE_GROUP_VCF);
988  43 sf.setSource(sourceId);
989   
990  43 sf.setValue(Gff3Helper.ALLELES, alleles);
991   
992  43 addAlleleProperties(variant, sf, altAlleleIndex, consequence);
993   
994  43 seq.addSequenceFeature(sf);
995   
996  43 return 1;
997    }
998   
999    /**
1000    * Determines the Sequence Ontology term to use for the variant feature type in
1001    * Jalview. The default is 'sequence_variant', but a more specific term is used
1002    * if:
1003    * <ul>
1004    * <li>VEP (or SnpEff) Consequence annotation is included in the VCF</li>
1005    * <li>sequence id can be matched to VEP Feature (or SnpEff Feature_ID)</li>
1006    * </ul>
1007    *
1008    * @param consequence
1009    * @return
1010    * @see http://www.sequenceontology.org/browser/current_svn/term/SO:0001060
1011    */
 
1012  7 toggle String getOntologyTerm(String consequence)
1013    {
1014  7 String type = SequenceOntologyI.SEQUENCE_VARIANT;
1015   
1016    /*
1017    * could we associate Consequence data with this allele and feature (transcript)?
1018    * if so, prefer the consequence term from that data
1019    */
1020  7 if (csqAlleleFieldIndex == -1) // && snpEffAlleleFieldIndex == -1
1021    {
1022    /*
1023    * no Consequence data so we can't refine the ontology term
1024    */
1025  0 return type;
1026    }
1027   
1028  7 if (consequence != null)
1029    {
1030  7 String[] csqFields = consequence.split(PIPE_REGEX);
1031  7 if (csqFields.length > csqConsequenceFieldIndex)
1032    {
1033  7 type = csqFields[csqConsequenceFieldIndex];
1034    }
1035    }
1036    else
1037    {
1038    // todo the same for SnpEff consequence data matching if wanted
1039    }
1040   
1041    /*
1042    * if of the form (e.g.) missense_variant&splice_region_variant,
1043    * just take the first ('most severe') consequence
1044    */
1045  7 if (type != null)
1046    {
1047  7 int pos = type.indexOf('&');
1048  7 if (pos > 0)
1049    {
1050  0 type = type.substring(0, pos);
1051    }
1052    }
1053  7 return type;
1054    }
1055   
1056    /**
1057    * Returns matched consequence data if it can be found, else null.
1058    * <ul>
1059    * <li>inspects the VCF data for key 'vcfInfoId'</li>
1060    * <li>splits this on comma (to distinct consequences)</li>
1061    * <li>returns the first consequence (if any) where</li>
1062    * <ul>
1063    * <li>the allele matches the altAlleleIndex'th allele of variant</li>
1064    * <li>the feature matches the sequence name (e.g. transcript id)</li>
1065    * </ul>
1066    * </ul>
1067    * If matched, the consequence is returned (as pipe-delimited fields).
1068    *
1069    * @param variant
1070    * @param vcfInfoId
1071    * @param altAlleleIndex
1072    * @param alleleFieldIndex
1073    * @param alleleNumberFieldIndex
1074    * @param seqName
1075    * @param featureFieldIndex
1076    * @return
1077    */
 
1078  43 toggle private String getConsequenceForAlleleAndFeature(VariantContext variant,
1079    String vcfInfoId, int altAlleleIndex, int alleleFieldIndex,
1080    int alleleNumberFieldIndex,
1081    String seqName, int featureFieldIndex)
1082    {
1083  43 if (alleleFieldIndex == -1 || featureFieldIndex == -1)
1084    {
1085  29 return null;
1086    }
1087  14 Object value = variant.getAttribute(vcfInfoId);
1088   
1089  14 if (value == null || !(value instanceof List<?>))
1090    {
1091  0 return null;
1092    }
1093   
1094    /*
1095    * inspect each consequence in turn (comma-separated blocks
1096    * extracted by htsjdk)
1097    */
1098  14 List<String> consequences = (List<String>) value;
1099   
1100  14 for (String consequence : consequences)
1101    {
1102  41 String[] csqFields = consequence.split(PIPE_REGEX);
1103  41 if (csqFields.length > featureFieldIndex)
1104    {
1105  41 String featureIdentifier = csqFields[featureFieldIndex];
1106  41 if (featureIdentifier.length() > 4
1107    && seqName.indexOf(featureIdentifier.toLowerCase()) > -1)
1108    {
1109    /*
1110    * feature (transcript) matched - now check for allele match
1111    */
1112  10 if (matchAllele(variant, altAlleleIndex, csqFields,
1113    alleleFieldIndex, alleleNumberFieldIndex))
1114    {
1115  7 return consequence;
1116    }
1117    }
1118    }
1119    }
1120  7 return null;
1121    }
1122   
 
1123  10 toggle private boolean matchAllele(VariantContext variant, int altAlleleIndex,
1124    String[] csqFields, int alleleFieldIndex,
1125    int alleleNumberFieldIndex)
1126    {
1127    /*
1128    * if ALLELE_NUM is present, it must match altAlleleIndex
1129    * NB first alternate allele is 1 for ALLELE_NUM, 0 for altAlleleIndex
1130    */
1131  10 if (alleleNumberFieldIndex > -1)
1132    {
1133  0 if (csqFields.length <= alleleNumberFieldIndex)
1134    {
1135  0 return false;
1136    }
1137  0 String alleleNum = csqFields[alleleNumberFieldIndex];
1138  0 return String.valueOf(altAlleleIndex + 1).equals(alleleNum);
1139    }
1140   
1141    /*
1142    * else consequence allele must match variant allele
1143    */
1144  10 if (alleleFieldIndex > -1 && csqFields.length > alleleFieldIndex)
1145    {
1146  10 String csqAllele = csqFields[alleleFieldIndex];
1147  10 String vcfAllele = variant.getAlternateAllele(altAlleleIndex)
1148    .getBaseString();
1149  10 return csqAllele.equals(vcfAllele);
1150    }
1151  0 return false;
1152    }
1153   
1154    /**
1155    * Add any allele-specific VCF key-value data to the sequence feature
1156    *
1157    * @param variant
1158    * @param sf
1159    * @param altAlelleIndex
1160    * (0, 1..)
1161    * @param consequence
1162    * if not null, the consequence specific to this sequence (transcript
1163    * feature) and allele
1164    */
 
1165  43 toggle protected void addAlleleProperties(VariantContext variant,
1166    SequenceFeature sf, final int altAlelleIndex, String consequence)
1167    {
1168  43 Map<String, Object> atts = variant.getAttributes();
1169   
1170  43 for (Entry<String, Object> att : atts.entrySet())
1171    {
1172  142 String key = att.getKey();
1173   
1174    /*
1175    * extract Consequence data (if present) that we are able to
1176    * associated with the allele for this variant feature
1177    */
1178  142 if (CSQ_FIELD.equals(key))
1179    {
1180  14 addConsequences(variant, sf, consequence);
1181  14 continue;
1182    }
1183   
1184    /*
1185    * filter out fields we don't want to capture
1186    */
1187  128 if (!vcfFieldsOfInterest.contains(key))
1188    {
1189  38 continue;
1190    }
1191   
1192    /*
1193    * filter out fields we don't want to capture
1194    */
1195  90 if (!vcfFieldsOfInterest.contains(key))
1196    {
1197  0 continue;
1198    }
1199   
1200    /*
1201    * we extract values for other data which are allele-specific;
1202    * these may be per alternate allele (INFO[key].Number = 'A')
1203    * or per allele including reference (INFO[key].Number = 'R')
1204    */
1205  90 VCFInfoHeaderLine infoHeader = header.getInfoHeaderLine(key);
1206  90 if (infoHeader == null)
1207    {
1208    /*
1209    * can't be sure what data belongs to this allele, so
1210    * play safe and don't take any
1211    */
1212  0 continue;
1213    }
1214   
1215  90 VCFHeaderLineCount number = infoHeader.getCountType();
1216  90 int index = altAlelleIndex;
1217  90 if (number == VCFHeaderLineCount.R)
1218    {
1219    /*
1220    * one value per allele including reference, so bump index
1221    * e.g. the 3rd value is for the 2nd alternate allele
1222    */
1223  14 index++;
1224    }
1225  76 else if (number != VCFHeaderLineCount.A)
1226    {
1227    /*
1228    * don't save other values as not allele-related
1229    */
1230  14 continue;
1231    }
1232   
1233    /*
1234    * take the index'th value
1235    */
1236  76 String value = getAttributeValue(variant, key, index);
1237  76 if (value != null)
1238    {
1239  76 sf.setValue(key, value);
1240    }
1241    }
1242    }
1243   
1244    /**
1245    * Inspects CSQ data blocks (consequences) and adds attributes on the sequence
1246    * feature.
1247    * <p>
1248    * If <code>myConsequence</code> is not null, then this is the specific
1249    * consequence data (pipe-delimited fields) that is for the current allele and
1250    * transcript (sequence) being processed)
1251    *
1252    * @param variant
1253    * @param sf
1254    * @param myConsequence
1255    */
 
1256  14 toggle protected void addConsequences(VariantContext variant, SequenceFeature sf,
1257    String myConsequence)
1258    {
1259  14 Object value = variant.getAttribute(CSQ_FIELD);
1260   
1261  14 if (value == null || !(value instanceof List<?>))
1262    {
1263  0 return;
1264    }
1265   
1266  14 List<String> consequences = (List<String>) value;
1267   
1268    /*
1269    * inspect CSQ consequences; restrict to the consequence
1270    * associated with the current transcript (Feature)
1271    */
1272  14 Map<String, String> csqValues = new HashMap<>();
1273   
1274  14 for (String consequence : consequences)
1275    {
1276  50 if (myConsequence == null || myConsequence.equals(consequence))
1277    {
1278  31 String[] csqFields = consequence.split(PIPE_REGEX);
1279   
1280    /*
1281    * inspect individual fields of this consequence, copying non-null
1282    * values which are 'fields of interest'
1283    */
1284  31 int i = 0;
1285  31 for (String field : csqFields)
1286    {
1287  279 if (field != null && field.length() > 0)
1288    {
1289  279 String id = vepFieldsOfInterest.get(i);
1290  279 if (id != null)
1291    {
1292  279 csqValues.put(id, field);
1293    }
1294    }
1295  279 i++;
1296    }
1297    }
1298    }
1299   
1300  14 if (!csqValues.isEmpty())
1301    {
1302  14 sf.setValue(CSQ_FIELD, csqValues);
1303    }
1304    }
1305   
1306    /**
1307    * A convenience method to complement a dna base and return the string value
1308    * of its complement
1309    *
1310    * @param reference
1311    * @return
1312    */
 
1313  0 toggle protected String complement(byte[] reference)
1314    {
1315  0 return String.valueOf(Dna.getComplement((char) reference[0]));
1316    }
1317   
1318    /**
1319    * Determines the location of the query range (chromosome positions) in a
1320    * different reference assembly.
1321    * <p>
1322    * If the range is just a subregion of one for which we already have a mapping
1323    * (for example, an exon sub-region of a gene), then the mapping is just
1324    * computed arithmetically.
1325    * <p>
1326    * Otherwise, calls the Ensembl REST service that maps from one assembly
1327    * reference's coordinates to another's
1328    *
1329    * @param queryRange
1330    * start-end chromosomal range in 'fromRef' coordinates
1331    * @param chromosome
1332    * @param species
1333    * @param fromRef
1334    * assembly reference for the query coordinates
1335    * @param toRef
1336    * assembly reference we wish to translate to
1337    * @return the start-end range in 'toRef' coordinates
1338    */
 
1339  0 toggle protected int[] mapReferenceRange(int[] queryRange, String chromosome,
1340    String species, String fromRef, String toRef)
1341    {
1342    /*
1343    * first try shorcut of computing the mapping as a subregion of one
1344    * we already have (e.g. for an exon, if we have the gene mapping)
1345    */
1346  0 int[] mappedRange = findSubsumedRangeMapping(queryRange, chromosome,
1347    species, fromRef, toRef);
1348  0 if (mappedRange != null)
1349    {
1350  0 return mappedRange;
1351    }
1352   
1353    /*
1354    * call (e.g.) http://rest.ensembl.org/map/human/GRCh38/17:45051610..45109016:1/GRCh37
1355    */
1356  0 EnsemblMap mapper = new EnsemblMap();
1357  0 int[] mapping = mapper.getAssemblyMapping(species, chromosome, fromRef,
1358    toRef, queryRange);
1359   
1360  0 if (mapping == null)
1361    {
1362    // mapping service failure
1363  0 return null;
1364    }
1365   
1366    /*
1367    * save mapping for possible future re-use
1368    */
1369  0 String key = makeRangesKey(chromosome, species, fromRef, toRef);
1370  0 if (!assemblyMappings.containsKey(key))
1371    {
1372  0 assemblyMappings.put(key, new HashMap<int[], int[]>());
1373    }
1374   
1375  0 assemblyMappings.get(key).put(queryRange, mapping);
1376   
1377  0 return mapping;
1378    }
1379   
1380    /**
1381    * If we already have a 1:1 contiguous mapping which subsumes the given query
1382    * range, this method just calculates and returns the subset of that mapping,
1383    * else it returns null. In practical terms, if a gene has a contiguous
1384    * mapping between (for example) GRCh37 and GRCh38, then we assume that its
1385    * subsidiary exons occupy unchanged relative positions, and just compute
1386    * these as offsets, rather than do another lookup of the mapping.
1387    * <p>
1388    * If in future these assumptions prove invalid (e.g. for bacterial dna?!),
1389    * simply remove this method or let it always return null.
1390    * <p>
1391    * Warning: many rapid calls to the /map service map result in a 429 overload
1392    * error response
1393    *
1394    * @param queryRange
1395    * @param chromosome
1396    * @param species
1397    * @param fromRef
1398    * @param toRef
1399    * @return
1400    */
 
1401  0 toggle protected int[] findSubsumedRangeMapping(int[] queryRange, String chromosome,
1402    String species, String fromRef, String toRef)
1403    {
1404  0 String key = makeRangesKey(chromosome, species, fromRef, toRef);
1405  0 if (assemblyMappings.containsKey(key))
1406    {
1407  0 Map<int[], int[]> mappedRanges = assemblyMappings.get(key);
1408  0 for (Entry<int[], int[]> mappedRange : mappedRanges.entrySet())
1409    {
1410  0 int[] fromRange = mappedRange.getKey();
1411  0 int[] toRange = mappedRange.getValue();
1412  0 if (fromRange[1] - fromRange[0] == toRange[1] - toRange[0])
1413    {
1414    /*
1415    * mapping is 1:1 in length, so we trust it to have no discontinuities
1416    */
1417  0 if (MappingUtils.rangeContains(fromRange, queryRange))
1418    {
1419    /*
1420    * fromRange subsumes our query range
1421    */
1422  0 int offset = queryRange[0] - fromRange[0];
1423  0 int mappedRangeFrom = toRange[0] + offset;
1424  0 int mappedRangeTo = mappedRangeFrom + (queryRange[1] - queryRange[0]);
1425  0 return new int[] { mappedRangeFrom, mappedRangeTo };
1426    }
1427    }
1428    }
1429    }
1430  0 return null;
1431    }
1432   
1433    /**
1434    * Transfers the sequence feature to the target sequence, locating its start
1435    * and end range based on the mapping. Features which do not overlap the
1436    * target sequence are ignored.
1437    *
1438    * @param sf
1439    * @param targetSequence
1440    * @param mapping
1441    * mapping from the feature's coordinates to the target sequence
1442    */
 
1443  0 toggle protected void transferFeature(SequenceFeature sf,
1444    SequenceI targetSequence, MapList mapping)
1445    {
1446  0 int[] mappedRange = mapping.locateInTo(sf.getBegin(), sf.getEnd());
1447   
1448  0 if (mappedRange != null)
1449    {
1450  0 String group = sf.getFeatureGroup();
1451  0 int newBegin = Math.min(mappedRange[0], mappedRange[1]);
1452  0 int newEnd = Math.max(mappedRange[0], mappedRange[1]);
1453  0 SequenceFeature copy = new SequenceFeature(sf, newBegin, newEnd,
1454    group, sf.getScore());
1455  0 targetSequence.addSequenceFeature(copy);
1456    }
1457    }
1458   
1459    /**
1460    * Formats a ranges map lookup key
1461    *
1462    * @param chromosome
1463    * @param species
1464    * @param fromRef
1465    * @param toRef
1466    * @return
1467    */
 
1468  0 toggle protected static String makeRangesKey(String chromosome, String species,
1469    String fromRef, String toRef)
1470    {
1471  0 return species + EXCL + chromosome + EXCL + fromRef + EXCL
1472    + toRef;
1473    }
1474    }