Support reference genome

Add GRCh38 example maf file

Update README.md

Author: Hongxin

.editorconfig

@@ -0,0 +1,22 @@
+# The EditorConfig project consists of a file format for defining coding styles
+# and a collection of text editor plugins that enable editors to read the file format
+# and adhere to defined styles.
+
+# EditorConfig files are read top to bottom and the closest EditorConfig files are read last.
+# Properties from matching EditorConfig sections are applied in the order they were read,
+# so properties in closer files take precedence.
+
+# Please only specify the formats you want to apply through out the project in this file.
+# Otherwise, please create new config file in your directory where you want to apply these styles.
+
+# More details about EditorConfig: http://EditorConfig.org
+
+# top-most EditorConfig file
+root = true
+
+[*]
+# Unix-style newlines with a newline ending every file
+insert_final_newline = false
+trim_trailing_whitespace = false
+
+

AnnotatorCore.py

@@ -114,6 +114,7 @@ def setsampleidsfileterfile(f):
 PROTEIN_POSITION_HEADERS = ['PROTEIN_POSITION']
 CANCER_TYPE_HEADERS = ['ONCOTREE_CODE', 'CANCER_TYPE']
 FUSION_HEADERS = ['FUSION']
+REFERENCE_GENOME_HEADERS = ['NCBI_BUILD', 'REFERENCE_GENOME']
 
 # columns for genomic change annotation
 GC_CHROMOSOME_HEADER = 'CHROMOSOME'
@@ -132,6 +133,11 @@ class QueryType(Enum):
     GENOMIC_CHANGE = 'GENOMIC_CHANGE'
 
 
+class ReferenceGenome(Enum):
+    GRCH37 = 'GRCh37'
+    GRCH38 = 'GRCh38'
+
+
 REQUIRED_QUERY_TYPE_COLUMNS = {
     QueryType.HGVSP_SHORT: [HGVSP_SHORT_HEADER],
     QueryType.HGVSP: [HGVSP_HEADER],
@@ -336,8 +342,19 @@ def resolve_query_type(user_input_query_type, headers):
     return selected_query_type
 
 
+def get_reference_genome_from_row(row_reference_genome, default_reference_genome):
+    reference_genome = default_reference_genome
+    if row_reference_genome is not None and row_reference_genome != '':
+        try:
+            reference_genome = ReferenceGenome[row_reference_genome.upper()]
+        except KeyError:
+            log.warning('Unexpected reference genome, only GRCh37 and GRCh38 are supported.' + (
+                ' Use default.' if default_reference_genome is not None else ' Skipping.'))
+    return reference_genome
+
+
 def processalterationevents(eventfile, outfile, previousoutfile, defaultCancerType, cancerTypeMap,
-                            retainonlycuratedgenes, annotatehotspots, user_input_query_type):
+                            retainonlycuratedgenes, annotatehotspots, user_input_query_type, default_reference_genome):
     if annotatehotspots:
         inithotspots()
     if os.path.isfile(previousoutfile):
@@ -381,19 +398,19 @@ def processalterationevents(eventfile, outfile, previousoutfile, defaultCancerTy
             process_alteration(reader, outf, headers, [HGVSP_SHORT_HEADER, ALTERATION_HEADER], ncols, newncols,
                                defaultCancerType,
                                cancerTypeMap,
-                               retainonlycuratedgenes, annotatehotspots)
+                               retainonlycuratedgenes, annotatehotspots, default_reference_genome)
 
         if (query_type == QueryType.HGVSP):
             process_alteration(reader, outf, headers, [HGVSP_HEADER, ALTERATION_HEADER], ncols, newncols, defaultCancerType,
                                cancerTypeMap,
-                               retainonlycuratedgenes, annotatehotspots)
+                               retainonlycuratedgenes, annotatehotspots, default_reference_genome)
 
         if (query_type == QueryType.HGVSG):
             process_hvsg(reader, outf, headers, [HGVSG_HEADER, ALTERATION_HEADER], ncols, newncols, defaultCancerType,
-                         cancerTypeMap)
+                         cancerTypeMap, default_reference_genome)
 
         if (query_type == QueryType.GENOMIC_CHANGE):
-            process_genomic_change(reader, outf, headers, ncols, newncols, defaultCancerType, cancerTypeMap)
+            process_genomic_change(reader, outf, headers, ncols, newncols, defaultCancerType, cancerTypeMap, default_reference_genome)
 
     outf.close()
 
@@ -407,7 +424,7 @@ def get_cell_content(row, index, return_empty_string=False):
         return None
 
 def process_alteration(maffilereader, outf, maf_headers, alteration_column_names, ncols, nannotationcols, defaultCancerType, cancerTypeMap,
-                       retainonlycuratedgenes, annotatehotspots):
+                       retainonlycuratedgenes, annotatehotspots, default_reference_genome):
     ihugo = geIndexOfHeader(maf_headers, HUGO_HEADERS)
     iconsequence = geIndexOfHeader(maf_headers, CONSEQUENCE_HEADERS)
     ihgvs = geIndexOfHeader(maf_headers, alteration_column_names)
@@ -416,6 +433,7 @@ def process_alteration(maffilereader, outf, maf_headers, alteration_column_names
     iend = geIndexOfHeader(maf_headers, PROTEIN_END_HEADERS)
     iproteinpos = geIndexOfHeader(maf_headers, PROTEIN_POSITION_HEADERS)
     icancertype = geIndexOfHeader(maf_headers, CANCER_TYPE_HEADERS)
+    ireferencegenome= geIndexOfHeader(maf_headers, REFERENCE_GENOME_HEADERS)
 
     posp = re.compile('[0-9]+')
 
@@ -445,6 +463,7 @@ def process_alteration(maffilereader, outf, maf_headers, alteration_column_names
             hgvs = hgvs[2:]
 
         cancertype = get_tumor_type_from_row(row, i, defaultCancerType, icancertype, cancerTypeMap, sample)
+        reference_genome = get_reference_genome_from_row(get_cell_content(row, ireferencegenome), default_reference_genome)
 
         hgvs = conversion(hgvs)
 
@@ -478,7 +497,7 @@ def process_alteration(maffilereader, outf, maf_headers, alteration_column_names
             row.append(_3dhotspot)
 
         if not retainonlycuratedgenes or hugo in curatedgenes:
-            query = ProteinChangeQuery(hugo, hgvs, cancertype, consequence, start, end)
+            query = ProteinChangeQuery(hugo, hgvs, cancertype, reference_genome, consequence, start, end)
             queries.append(query)
             rows.append(row)
         else:
@@ -510,7 +529,7 @@ def get_var_allele(ref_allele, tumor_seq_allele1, tumor_seq_allele2):
 
     return tumor_seq_allele
 
-def process_genomic_change(maffilereader, outf, maf_headers, ncols, nannotationcols, defaultCancerType, cancerTypeMap):
+def process_genomic_change(maffilereader, outf, maf_headers, ncols, nannotationcols, defaultCancerType, cancerTypeMap, default_reference_genome):
     ichromosome = geIndexOfHeader(maf_headers, [GC_CHROMOSOME_HEADER])
     istart = geIndexOfHeader(maf_headers, [GC_START_POSITION_HEADER])
     iend = geIndexOfHeader(maf_headers, [GC_END_POSITION_HEADER])
@@ -520,6 +539,7 @@ def process_genomic_change(maffilereader, outf, maf_headers, ncols, nannotationc
 
     isample = geIndexOfHeader(maf_headers, SAMPLE_HEADERS)
     icancertype = geIndexOfHeader(maf_headers, CANCER_TYPE_HEADERS)
+    ireferencegenome= geIndexOfHeader(maf_headers, REFERENCE_GENOME_HEADERS)
 
     posp = re.compile('[0-9]+')
 
@@ -539,6 +559,7 @@ def process_genomic_change(maffilereader, outf, maf_headers, ncols, nannotationc
             continue
 
         cancertype = get_tumor_type_from_row(row, i, defaultCancerType, icancertype, cancerTypeMap, sample)
+        reference_genome = get_reference_genome_from_row(get_cell_content(row, ireferencegenome), default_reference_genome)
 
         chromosome = get_cell_content(row, ichromosome, True)
         start = get_cell_content(row, istart, True)
@@ -548,7 +569,7 @@ def process_genomic_change(maffilereader, outf, maf_headers, ncols, nannotationc
         var_allele_2 = get_cell_content(row, ivarallele2, True)
         var_allele = get_var_allele(ref_allele, var_allele_1, var_allele_2)
 
-        query = GenomicChangeQuery(chromosome, start, end, ref_allele, var_allele, cancertype)
+        query = GenomicChangeQuery(chromosome, start, end, ref_allele, var_allele, cancertype, reference_genome)
         queries.append(query)
         rows.append(row)
 
@@ -562,10 +583,11 @@ def process_genomic_change(maffilereader, outf, maf_headers, ncols, nannotationc
         annotations = pull_genomic_change_info(queries)
         append_annotation_to_file(outf, ncols+nannotationcols, rows, annotations)
 
-def process_hvsg(maffilereader, outf, maf_headers, alteration_column_names, ncols, nannotationcols, defaultCancerType, cancerTypeMap):
+def process_hvsg(maffilereader, outf, maf_headers, alteration_column_names, ncols, nannotationcols, defaultCancerType, cancerTypeMap, default_reference_genome):
     ihgvsg = geIndexOfHeader(maf_headers, alteration_column_names)
     isample = geIndexOfHeader(maf_headers, SAMPLE_HEADERS)
     icancertype = geIndexOfHeader(maf_headers, CANCER_TYPE_HEADERS)
+    ireferencegenome= geIndexOfHeader(maf_headers, REFERENCE_GENOME_HEADERS)
 
     i = 0
     queries = []
@@ -585,12 +607,13 @@ def process_hvsg(maffilereader, outf, maf_headers, alteration_column_names, ncol
         hgvsg = get_cell_content(row, ihgvsg)
 
         cancertype = get_tumor_type_from_row(row, i, defaultCancerType, icancertype, cancerTypeMap, sample)
+        reference_genome = get_reference_genome_from_row(get_cell_content(row, ireferencegenome), default_reference_genome)
 
         if hgvsg is None:
             append_annotation_to_file(outf, ncols + nannotationcols, [row],
                                       [[GENE_IN_ONCOKB_DEFAULT, VARIANT_IN_ONCOKB_DEFAULT]])
         else:
-            query = HGVSgQuery(hgvsg, cancertype)
+            query = HGVSgQuery(hgvsg, cancertype, reference_genome)
             queries.append(query)
             rows.append(row)
 
@@ -1201,7 +1224,7 @@ def __init__(self, hugo):
 
 
 class ProteinChangeQuery:
-    def __init__(self, hugo, hgvs, cancertype, consequence=None, start=None, end=None):
+    def __init__(self, hugo, hgvs, cancertype, reference_genome=None, consequence=None, start=None, end=None):
         self.gene = Gene(hugo)
         self.alteration = hgvs
         if consequence is not None:
@@ -1211,16 +1234,24 @@ def __init__(self, hugo, hgvs, cancertype, consequence=None, start=None, end=Non
         if end is not None:
             self.proteinEnd = end
         self.tumorType = cancertype
+        if reference_genome is not None:
+            self.referenceGenome = reference_genome.value
+
 
 class HGVSgQuery:
-    def __init__(self, hgvsg, cancertype):
+    def __init__(self, hgvsg, cancertype, reference_genome=None):
         self.hgvsg = hgvsg
         self.tumorType = cancertype
+        if reference_genome is not None:
+            self.referenceGenome = reference_genome.value
+
 
 class GenomicChangeQuery:
-    def __init__(self, chromosome, start, end, ref_allele, var_allele, cancertype):
+    def __init__(self, chromosome, start, end, ref_allele, var_allele, cancertype, reference_genome=None):
         self.genomicLocation = ','.join([chromosome, start, end, ref_allele, var_allele])
         self.tumorType = cancertype
+        if reference_genome is not None:
+            self.referenceGenome = reference_genome.value
 
 class CNAQuery:
     def __init__(self, hugo, cnatype, cancertype):

MafAnnotator.py

@@ -13,7 +13,7 @@ def main(argv):
         log.info(
             '\n'
             'MafAnnotator.py -i <input MAF file> -o <output MAF file> [-p previous results] [-c <input clinical file>] '
-            '[-s sample list filter] [-t <default tumor type>] [-u oncokb-base-url] [-b oncokb api bear token] [-a] [-q query type]\n'
+            '[-s sample list filter] [-t <default tumor type>] [-u oncokb-base-url] [-b oncokb api bear token] [-a] [-q query type] [-r defauult reference genome]\n'
             'Essential MAF columns (case insensitive):\n'
             '    HUGO_SYMBOL: Hugo gene symbol\n'
             '    VARIANT_CLASSIFICATION: Translational effect of variant allele\n'
@@ -30,14 +30,17 @@ def main(argv):
             '    2) ONCOTREE_CODE exist in MAF\n'
             '    3) default tumor type (-t)\n'
             'Query type only allows the following values (case-insensitive):\n'
-            '    - HGVSp_Short \n'
+            '    - HGVSp_Short\n'
             '      It reads from column HGVSp_Short or Alteration\n'
             '    - HGVSp\n'
             '      It reads from column HGVSp or Alteration\n'
             '    - HGVSg\n'
             '      It reads from column HGVSg or Alteration\n'
             '    - Genomic_Change\n'
             '      It reads from columns Chromosome, Start_Position, End_Position, Reference_Allele, Tumor_Seq_Allele1 and Tumor_Seq_Allele2  \n'
+            'Reference Genome only allows the following values(case-insensitive):\n'
+            '    - GRCh37\n'
+            '      GRCh38\n'
             'Default OncoKB base url is https://www.oncokb.org.\n'
             'Use -a to annotate mutational hotspots\n'
         )
@@ -66,13 +69,21 @@ def main(argv):
         try:
             user_input_query_type = QueryType[argv.query_type.upper()]
         except KeyError:
-            # if not isinstance(argv.query_type.upper(), QueryType):
-            print(
+            log.error(
                 'Query type is not acceptable. Only the following allows(case insensitive): HGVSp_Short, HGVSp, HGVSg, Genomic_Change')
             raise
 
+    default_reference_genome = None
+    if argv.default_reference_genome is not None:
+        try:
+            default_reference_genome = ReferenceGenome[argv.default_reference_genome.upper()]
+        except KeyError:
+            log.error(
+                'Reference genome is not acceptable. Only the following allows(case insensitive): GRCh37, GRCh38')
+            raise
+
     processalterationevents(argv.input_file, argv.output_file, argv.previous_result_file, argv.default_cancer_type,
-                            cancertypemap, True, argv.annotate_hotspots, user_input_query_type)
+                            cancertypemap, True, argv.annotate_hotspots, user_input_query_type, default_reference_genome)
 
     log.info('done!')
 
@@ -91,6 +102,7 @@ def main(argv):
     parser.add_argument('-v', dest='cancer_hotspots_base_url', default='', type=str)
     parser.add_argument('-b', dest='oncokb_api_bearer_token', default='', type=str)
     parser.add_argument('-q', dest='query_type', default=None, type=str)
+    parser.add_argument('-r', dest='default_reference_genome', default=None, type=str)
     parser.set_defaults(func=main)
 
     args = parser.parse_args()

README.md

@@ -31,7 +31,7 @@ Example input files are under [data](data). An example script is here: [example.
 We recommend processing VCF files by [vcf2maf](https://github.com/mskcc/vcf2maf/) with [MSK override isoforms](https://github.com/mskcc/vcf2maf/blob/master/data/isoform_overrides_at_mskcc) before using the `MafAnnotator` here.
 
 
-#### Annotate with HGVSp_Short, HGVSp, HGVSg or Genomic Change
+### Annotate with HGVSp_Short, HGVSp, HGVSg or Genomic Change
 OncoKB MafAnnotator supports annotating the alteration with HGVSp, HGVSp_Short, HGVSg or Genomic Change format. Please specify the query type with -q parameter.
 The acceptable values are HGVSp_Short, HGVSp, HGVSg and Genomic_Change(case-insensitive). Please see data/example.sh for examples.  
 If you do not specify query type, the MafAnnotator will try to figure out the query type based on the headers.  
@@ -42,6 +42,18 @@ For HGVSg, the annotator takes alteration from the column HGVSg or Alteration
 For Genomic_Change, the annotator takes genomic change from columns Chromosome, Start_Position, End_Position, Reference_Allele, Tumor_Seq_Allele1 and Tumor_Seq_Allele2  
 
 
+### Annotate with different reference genomes (GRCh37, GRCh38)
+OncoKB MafAnnotator supports annotating the alteration with reference genome GRCh37 and GRCh38.  
+
+The annotator will get the reference genome from MAF file column NCBI_Build or Reference_Genome.  
+If there is no reference genome specified in the file, we will use the default reference genome through -r parameter.  
+
+You can specify the default reference genome using -r parameter (This is only applicable to MafAnnotator.py).  
+The acceptable values are GRCh37, GRCh38 (case in-sensitive).  
+
+If both values are not specified, the annotator will use OncoKB default reference genome which is GRCh37.
+
+
 ## Levels of Evidence
 Introducing [Simplified OncoKB Levels of Evidence](https://www.oncokb.org/levels):
 - New Level 2, defined as “Standard care biomarker recommended by the NCCN or other expert panels predictive of response to an FDA-approved drug in this indication” (formerly Level 2A).

data/example_maf.txt

@@ -1,16 +1,17 @@
-NCBI_Build	Hugo_Symbol	Variant_Classification	Tumor_Sample_Barcode	HGVSp_Short	HGVSp	HGVSg	Chromosome	Start_Position	End_Position	Reference_Allele	Tumor_Seq_Allele1	Tumor_Seq_Allele2
-GRCh37	CUL1	Missense_Mutation	TCGA-A6-2672-01A-01W-0833-10	p.Y466S	Tyr466Ser							
-GRCh37	AKT3	Nonsense_Mutation	TCGA-05-4417-01	p.E182*	Glu182*							
-GRCh37	PIK3CA	Missense_Mutation	TCGA-02-0033-01	p.E542K	Glu542Lys	3:g.178936082G>A	3	178936082	178936082	G	A	A
-GRCh37	FGFR3	Missense_Mutation	TCGA-05-4417-01	p.V271M	Val271Met							
-GRCh37	EGFR	Missense_Mutation	TCGA-06-0155-01	p.H304Y	His304Tyr	7:g.55223543C>T	7	55223543	55223543	C	T	T
-GRCh37	PTEN	Missense_Mutation	TCGA-06-0155-01	p.C136R	Cys136Arg	10:g.89692922T>C	10	89692922	89692922	T	C	C
-GRCh37	FGFR2	Missense_Mutation	TCGA-02-0033-01	p.Q212K	Gln121Lys							
-GRCh37	ATM	Missense_Mutation	TCGA-05-4417-01	p.L2890R	Leu2890Arg							
-GRCh37	KRAS	Missense_Mutation	TCGA-05-4417-01	p.G12C	Gly12Cys	12:g.25398285C>A	12	25398285	25398285	C	A	A
-GRCh37	KRAS	Missense_Mutation	TCGA-05-4417-01	p.G12C	Gly12Cys	12:g.25398285_25398286delinsAG	12	25398285	25398286	CA	AG	AG
-GRCh37	RB1	Nonsense_Mutation	TCGA-02-0033-01	p.Q702*	Gln702*							
-GRCh37	TP53	Missense_Mutation	TCGA-02-0033-01	p.R248Q	Arg248Gln	17:g.7577538C>T	17	7577538	7577538	C	T	T
-GRCh37	NF1	Splice_Site	TCGA-02-0033-01	p.X1445_splice	X1445_splice	17:g.29586049G>A	17	29586049	29586049	G	A	A
-GRCh37	STK11	Missense_Mutation	TCGA-05-4417-01	p.H168R	His168Arg							
-GRCh37	TERT	5'Flank	TCGA-05-4417-01			5:g.1295228G>A	5	1295228	1295228	G	A	A
+NCBI_Build	Hugo_Symbol	Variant_Classification	Tumor_Sample_Barcode	HGVSp_Short	HGVSp	HGVSg	Chromosome	Start_Position	End_Position	Reference_Allele	Tumor_Seq_Allele1	Tumor_Seq_Allele2
+GRCh37	CUL1	Missense_Mutation	TCGA-A6-2672-01A-01W-0833-10	p.Y466S	Tyr466Ser							
+GRCh37	AKT3	Nonsense_Mutation	TCGA-05-4417-01	p.E182*	Glu182*							
+GRCh37	PIK3CA	Missense_Mutation	TCGA-02-0033-01	p.E542K	Glu542Lys	3:g.178936082G>A	3	178936082	178936082	G	A	A
+GRCh37	FGFR3	Missense_Mutation	TCGA-05-4417-01	p.V271M	Val271Met							
+GRCh37	EGFR	Missense_Mutation	TCGA-06-0155-01	p.H304Y	His304Tyr	7:g.55223543C>T	7	55223543	55223543	C	T	T
+GRCh37	PTEN	Missense_Mutation	TCGA-06-0155-01	p.C136R	Cys136Arg	10:g.89692922T>C	10	89692922	89692922	T	C	C
+GRCh37	FGFR2	Missense_Mutation	TCGA-02-0033-01	p.Q212K	Gln121Lys							
+GRCh37	ATM	Missense_Mutation	TCGA-05-4417-01	p.L2890R	Leu2890Arg							
+GRCh37	KRAS	Missense_Mutation	TCGA-05-4417-01	p.G12C	Gly12Cys	12:g.25398285C>A	12	25398285	25398285	C	A	A
+GRCh37	KRAS	Missense_Mutation	TCGA-05-4417-01	p.G12C	Gly12Cys	12:g.25398285_25398286delinsAG	12	25398285	25398286	CA	AG	AG
+GRCh37	RB1	Nonsense_Mutation	TCGA-02-0033-01	p.Q702*	Gln702*							
+GRCh37	TP53	Missense_Mutation	TCGA-02-0033-01	p.R248Q	Arg248Gln	17:g.7577538C>T	17	7577538	7577538	C	T	T
+GRCh37	NF1	Splice_Site	TCGA-02-0033-01	p.X1445_splice	X1445_splice	17:g.29586049G>A	17	29586049	29586049	G	A	A
+GRCh37	STK11	Missense_Mutation	TCGA-05-4417-01	p.H168R	His168Arg							
+GRCh37	TERT	5'Flank	TCGA-05-4417-01			5:g.1295228G>A	5	1295228	1295228	G	A	A
+GRCh37	MYD88	Missense_Mutation	TCGA-05-4417-01	M232T