Release/0.3.0

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.2.14
+current_version = 0.2.17
 commit = True
 tag = False
 parse = (?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>[a-z0-9+]+)

.gitignore

@@ -5,3 +5,5 @@ __pycache__/
 build/
 venv/
 working/
+data/
+workspace/

CHANGELOG.md

@@ -2,6 +2,12 @@
 
 ## dev
 
+- [17](https://github.com/umccr/bolt/pull/17) - change dragen HRD file optional
+
+- [14](https://github.com/umccr/bolt/pull/14) - gpgr version bump to 2.2.0
+
 - [3](https://github.com/scwatts/bolt/pull/3) - Improve PCGR / CPSR argument handling
 
-- [6](https://github.com/umccr/bolt/pull/6) - Change oncoanalyser v2.0.0 uptade, with switch sv caller from GRIPSS to eSVee
+- [6](https://github.com/umccr/bolt/pull/6) - Change oncoanalyser v2.0.0 uptade, with switch sv caller from GRIPSS to eSVee
+
+- [9](https://github.com/umccr/bolt/pull/9) Add hypermutation sample handling

README.md

@@ -54,20 +54,20 @@ the software environment and dependencies. Consequently, dependencies are split
 
 | Name    | Docker image URI                    | Commands                       |
 | ---     | ---                                 | ---                            |
-| pcgr    | ghcr.io/scwatts/bolt:0.2.14-pcgr    | • `bolt smlv_germline report`<br />• `bolt smlv_somatic annotate`<br />• `bolt smlv_somatic report`<br /> |
-| gpgr    | ghcr.io/scwatts/bolt:0.2.14-gpgr    | • `bolt other cancer_report`   |
-| snpeff  | ghcr.io/scwatts/bolt:0.2.14-snpeff  | • `bolt sv_somatic annotate`   |
-| circos  | ghcr.io/scwatts/bolt:0.2.14-circos  | • `bolt other purple_baf_plot` |
-| multiqc | ghcr.io/scwatts/bolt:0.2.14-multiqc | • `bolt other multiqc_report`  |
-| base    | ghcr.io/scwatts/bolt:0.2.14         | • `bolt smlv_germline prepare`<br />• `bolt smlv_somatic rescue`<br />• `bolt smlv_somatic filter`<br />• `bolt sv_somatic prioritise`<br /> |
+| pcgr    | ghcr.io/scwatts/bolt:0.2.17-pcgr    | • `bolt smlv_germline report`<br />• `bolt smlv_somatic annotate`<br />• `bolt smlv_somatic report`<br /> |
+| gpgr    | ghcr.io/scwatts/bolt:0.2.17-gpgr    | • `bolt other cancer_report`   |
+| snpeff  | ghcr.io/scwatts/bolt:0.2.17-snpeff  | • `bolt sv_somatic annotate`   |
+| circos  | ghcr.io/scwatts/bolt:0.2.17-circos  | • `bolt other purple_baf_plot` |
+| multiqc | ghcr.io/scwatts/bolt:0.2.17-multiqc | • `bolt other multiqc_report`  |
+| base    | ghcr.io/scwatts/bolt:0.2.17         | • `bolt smlv_germline prepare`<br />• `bolt smlv_somatic rescue`<br />• `bolt smlv_somatic filter`<br />• `bolt sv_somatic prioritise`<br /> |
 
 ## Usage
 
 Given the nature of software dependencies required, it is strongly recommended to run `bolt` commands via the existing
 [Docker images](#docker-images):
 
 ```bash
-docker run -ti -v $(pwd):$(pwd) -w $(pwd) ghcr.io/scwatts/bolt:0.2.14 \
+docker run -ti -v $(pwd):$(pwd) -w $(pwd) ghcr.io/scwatts/bolt:0.2.17 \
   bolt smlv_somatic filter \
     --tumor_name tumor_sample \
     --vcf_fp tumor_sample.vcf.gz \

bolt/common/constants.py

@@ -4,7 +4,7 @@
 ######################################
 ## Variation selection (annotation) ##
 ######################################
-MAX_SOMATIC_VARIANTS = 500_000
+MAX_SOMATIC_VARIANTS = 450_000
 MAX_SOMATIC_VARIANTS_GNOMAD_FILTER = 0.01
 
 
@@ -35,10 +35,47 @@
     'pathogenic',
     'uncertain_significance',
 }
-PCGR_TIERS_RESCUE = {
+PCGR_ACTIONABILITY_TIER_RESCUE = {
+    '1',
+    '2',
+}
+
+
+################################
+## Hypermutated report filter ##
+################################
+PCGR_TIERS_FILTERING = (
     'TIER_1',
     'TIER_2',
-}
+    'TIER_3',
+    'TIER_4',
+    'NONCODING',
+)
+
+VEP_IMPACTS_FILTER = (
+    'intergenic',
+    'intronic',
+    'downstream',
+    'upstream',
+    'impacts_other',
+)
+
+GENOMIC_REGIONS_FILTERING = (
+    'difficult',
+    'none',
+    'giab_conf',
+)
+
+HOTSPOT_FIELDS_FILTERING = (
+    'SAGE_HOTSPOT',
+    'hotspot',
+    'PCGR_MUTATION_HOTSPOT',
+)
+
+RETAIN_FIELDS_FILTERING = (
+    'PANEL',
+    *HOTSPOT_FIELDS_FILTERING,
+)
 
 
 ##################################################
@@ -61,6 +98,8 @@ class VcfFilter(enum.Enum):
     ENCODE = 'ENCODE'
     GNOMAD_COMMON = 'gnomAD_common'
 
+    PCGR_COUNT_LIMIT = 'PCGR_count_limit'
+
     @property
     def namespace(self):
         return 'FILTER'
@@ -77,16 +116,14 @@ class VcfInfo(enum.Enum):
     SAGE_NOVEL = 'SAGE_NOVEL'
     SAGE_RESCUE = 'SAGE_RESCUE'
 
-    PCGR_TIER = 'PCGR_TIER'
+    PCGR_ACTIONABILITY_TIER = 'PCGR_ACTIONABILITY_TIER'
     PCGR_CSQ = 'PCGR_CSQ'
     PCGR_MUTATION_HOTSPOT = 'PCGR_MUTATION_HOTSPOT'
-    PCGR_CLINVAR_CLNSIG = 'PCGR_CLINVAR_CLNSIG'
+    PCGR_CLINVAR_CLASSIFICATION = 'PCGR_CLINVAR_CLASSIFICATION'
     PCGR_COSMIC_COUNT = 'PCGR_COSMIC_COUNT'
     PCGR_TCGA_PANCANCER_COUNT = 'PCGR_TCGA_PANCANCER_COUNT'
     PCGR_ICGC_PCAWG_COUNT = 'PCGR_ICGC_PCAWG_COUNT'
 
-    CPSR_FINAL_CLASSIFICATION = 'CPSR_FINAL_CLASSIFICATION'
-    CPSR_PATHOGENICITY_SCORE = 'CPSR_PATHOGENICITY_SCORE'
     CPSR_CLINVAR_CLASSIFICATION = 'CPSR_CLINVAR_CLASSIFICATION'
     CPSR_CSQ = 'CPSR_CSQ'
 
@@ -112,7 +149,7 @@ class VcfInfo(enum.Enum):
 
     GNOMAD_AF = 'gnomAD_AF'
 
-    PCGR_TIER_RESCUE = 'PCGR_TIER_RESCUE'
+    PCGR_ACTIONABILITY_TIER_RESCUE = 'PCGR_ACTIONABILITY_TIER_RESCUE'
     SAGE_HOTSPOT_RESCUE = 'SAGE_HOTSPOT_RESCUE'
     CLINICAL_POTENTIAL_RESCUE = 'CLINICAL_POTENTIAL_RESCUE'
 
@@ -121,6 +158,8 @@ class VcfInfo(enum.Enum):
     RESCUED_FILTERS_EXISTING = 'RESCUED_FILTERS_EXISTING'
     RESCUED_FILTERS_PENDING = 'RESCUED_FILTERS_PENDING'
 
+    PANEL = 'PANEL'
+
     @property
     def namespace(self):
         return 'INFO'
@@ -187,6 +226,9 @@ def namespace(self):
         'Description': f'gnomAD AF >= {MAX_GNOMAD_AF}',
     },
 
+    VcfFilter.PCGR_COUNT_LIMIT: {
+        'Description': 'Manually filtered to meet PCGR 500,000 variant limit',
+    },
 
     # INFO
     VcfInfo.TUMOR_AF: {
@@ -226,7 +268,7 @@ def namespace(self):
         'Description': 'Variant rescued by a matching SAGE call',
     },
 
-    VcfInfo.PCGR_TIER: {
+    VcfInfo.PCGR_ACTIONABILITY_TIER: {
         'Number': '1',
         'Type': 'String',
         'Description': (
@@ -237,28 +279,29 @@ def namespace(self):
     },
     VcfInfo.PCGR_CSQ: {
         'Number': '.',
-        'Type': 'String',
-        'Description': (
-            'Consequence annotations from Ensembl VEP. Format: Allele|Consequence|IMPACT|SYMBOL|'
-            'Gene|Feature_type|Feature|BIOTYPE|EXON|INTRON|HGVSc|HGVSp|cDNA_position|'
-            'CDS_position|Protein_position|Amino_acids|Codons|Existing_variation|ALLELE_NUM|'
-            'DISTANCE|STRAND|FLAGS|PICK|VARIANT_CLASS|SYMBOL_SOURCE|HGNC_ID|CANONICAL|'
-            'MANE_SELECT|MANE_PLUS_CLINICAL|TSL|APPRIS|CCDS|ENSP|SWISSPROT|TREMBL|UNIPARC|'
-            'UNIPROT_ISOFORM|RefSeq|DOMAINS|HGVS_OFFSET|AF|AFR_AF|AMR_AF|EAS_AF|EUR_AF|SAS_AF|'
-            'gnomAD_AF|gnomAD_AFR_AF|gnomAD_AMR_AF|gnomAD_ASJ_AF|gnomAD_EAS_AF|gnomAD_FIN_AF|'
-            'gnomAD_NFE_AF|gnomAD_OTH_AF|gnomAD_SAS_AF|CLIN_SIG|SOMATIC|PHENO|CHECK_REF|'
-            'NearestExonJB'
-        ),
+    'Type': 'String',
+    'Description': (
+        'Consequence annotations from Ensembl VEP. Format: '
+        'Allele|Consequence|IMPACT|SYMBOL|Gene|Feature_type|Feature|BIOTYPE|EXON|INTRON|HGVSc|'
+        'HGVSp|cDNA_position|CDS_position|Protein_position|Amino_acids|Codons|Existing_variation|'
+        'ALLELE_NUM|DISTANCE|STRAND|FLAGS|PICK|VARIANT_CLASS|SYMBOL_SOURCE|HGNC_ID|CANONICAL|'
+        'MANE|MANE_SELECT|MANE_PLUS_CLINICAL|TSL|APPRIS|CCDS|ENSP|SWISSPROT|TREMBL|UNIPARC|'
+        'UNIPROT_ISOFORM|RefSeq|DOMAINS|HGVS_OFFSET|gnomADe_AF|gnomADe_AFR_AF|gnomADe_AMR_AF|'
+        'gnomADe_ASJ_AF|gnomADe_EAS_AF|gnomADe_FIN_AF|gnomADe_MID_AF|gnomADe_NFE_AF|'
+        'gnomADe_REMAINING_AF|gnomADe_SAS_AF|CLIN_SIG|SOMATIC|PHENO|CHECK_REF|MOTIF_NAME|'
+        'MOTIF_POS|HIGH_INF_POS|MOTIF_SCORE_CHANGE|TRANSCRIPTION_FACTORS|NearestExonJB|'
+        'MaxEntScan_alt|MaxEntScan_diff|MaxEntScan_ref'
+    ),
     },
     VcfInfo.PCGR_MUTATION_HOTSPOT: {
         'Number': '.',
         'Type': 'String',
-        'Description': 'Known cancer mutation hotspot, as found in cancerhotspots.org_v2, Gene|Codon|Q-value',
+        'Description': 'Known cancer mutation hotspot, as found in cancerhotspots.org. Format: GeneSymbol|Entrez_ID|CodonRefAA|Alt_AA|Q-value',
     },
-    VcfInfo.PCGR_CLINVAR_CLNSIG: {
+    VcfInfo.PCGR_CLINVAR_CLASSIFICATION: {
         'Number': '.',
         'Type': 'String',
-        'Description': 'ClinVar clinical significance',
+        'Description': 'ClinVar - Overall clinical significance of variant on a five-tiered scale',
     },
     VcfInfo.PCGR_COSMIC_COUNT: {
         'Number': '1',
@@ -276,23 +319,10 @@ def namespace(self):
         'Description': 'Count of ICGC PCAWG hits',
     },
 
-    VcfInfo.CPSR_FINAL_CLASSIFICATION: {
-        'Number': '1',
-        'Type': 'String',
-        'Description': (
-            'Final variant classification based on the combination of CLINVAR_CLASSIFICTION (for '
-            'ClinVar-classified variants), and CPSR_CLASSIFICATION (for novel variants)'
-        ),
-    },
-    VcfInfo.CPSR_PATHOGENICITY_SCORE: {
-        'Number': '1',
-        'Type': 'Float',
-        'Description': 'Aggregated CPSR pathogenicity score',
-    },
     VcfInfo.CPSR_CLINVAR_CLASSIFICATION: {
         'Number': '1',
         'Type': 'String',
-        'Description': 'Clinical significance of variant on a five-tiered scale',
+        'Description': 'ClinVar - Overall clinical significance of variant on a five-tiered scale',
     },
     VcfInfo.CPSR_CSQ: {
         'Number': '.',
@@ -301,13 +331,13 @@ def namespace(self):
             'Consequence annotations from Ensembl VEP. Format: Allele|Consequence|IMPACT|SYMBOL|'
             'Gene|Feature_type|Feature|BIOTYPE|EXON|INTRON|HGVSc|HGVSp|cDNA_position|CDS_position|'
             'Protein_position|Amino_acids|Codons|Existing_variation|ALLELE_NUM|DISTANCE|STRAND|'
-            'FLAGS|PICK|VARIANT_CLASS|SYMBOL_SOURCE|HGNC_ID|CANONICAL|MANE_SELECT|'
-            'MANE_PLUS_CLINICAL|APPRIS|CCDS|ENSP|SWISSPROT|TREMBL|UNIPARC|UNIPROT_ISOFORM|RefSeq|'
-            'DOMAINS|HGVS_OFFSET|AF|AFR_AF|AMR_AF|EAS_AF|EUR_AF|SAS_AF|gnomAD_AF|gnomAD_AFR_AF|'
-            'gnomAD_AMR_AF|gnomAD_ASJ_AF|gnomAD_EAS_AF|gnomAD_FIN_AF|gnomAD_NFE_AF|gnomAD_OTH_AF|'
-            'gnomAD_SAS_AF|CLIN_SIG|SOMATIC|PHENO|CHECK_REF|MOTIF_NAME|MOTIF_POS|HIGH_INF_POS|'
-            'MOTIF_SCORE_CHANGE|TRANSCRIPTION_FACTORS|NearestExonJB|LoF|LoF_filter|LoF_flags|'
-            'LoF_info'
+            'FLAGS|PICK|VARIANT_CLASS|SYMBOL_SOURCE|HGNC_ID|CANONICAL|MANE|MANE_SELECT|'
+            'MANE_PLUS_CLINICAL|TSL|APPRIS|CCDS|ENSP|SWISSPROT|TREMBL|UNIPARC|UNIPROT_ISOFORM|RefSeq|'
+            'DOMAINS|HGVS_OFFSET|gnomADe_AF|gnomADe_AFR_AF|gnomADe_AMR_AF|gnomADe_ASJ_AF|'
+            'gnomADe_EAS_AF|gnomADe_FIN_AF|gnomADe_MID_AF|gnomADe_NFE_AF|gnomADe_REMAINING_AF|'
+            'gnomADe_SAS_AF|CLIN_SIG|SOMATIC|PHENO|CHECK_REF|MOTIF_NAME|MOTIF_POS|HIGH_INF_POS|'
+            'MOTIF_SCORE_CHANGE|TRANSCRIPTION_FACTORS|NearestExonJB|MaxEntScan_alt|MaxEntScan_diff|'
+            'MaxEntScan_ref'
         ),
     },
 
@@ -316,7 +346,7 @@ def namespace(self):
         'Type': 'Flag',
         'Description': '',
     },
-    VcfInfo.PCGR_TIER_RESCUE: {
+    VcfInfo.PCGR_ACTIONABILITY_TIER_RESCUE: {
         'Number': '0',
         'Type': 'Flag',
         'Description': '',
@@ -350,6 +380,12 @@ def namespace(self):
         'Description': 'Filters pending prior to variant rescue',
     },
 
+    VcfInfo.PANEL: {
+        'Number': '0',
+        'Type': 'Flag',
+        'Description': 'UMCCR somatic panel CDS (2,000 bp padding)',
+    },
+
 
     # FORMAT
     VcfFormat.SAGE_AD: {