update oncogenicity classification

pcgr/annoutils.py

@@ -303,14 +303,10 @@ def assign_cds_exon_intron_annotations(csq_record, grantham_scores, logger):
                     if '-' in cds_pos_full and not '?' in cds_pos_full:
                         cds_pos = cds_pos_full.split('-')[0]
                         if cds_pos.isdigit():
-                            cds_pos = int(cds_pos)
-                        #else:
-                        #    logger.warning(f'Could not determine variant CDS position from VEP annotation - ({csq_record["CDS_position"]})')                        
+                            cds_pos = int(cds_pos)                   
                     else:
                         if cds_pos_full.isdigit():
-                            cds_pos = int(cds_pos_full)
-                        #else:
-                        #    logger.warning(f'Could not determine variant CDS position from VEP annotation - ({csq_record["CDS_position"]})')                         
+                            cds_pos = int(cds_pos_full)                     
 
                     if int(cds_pos) > -1 and int(cds_pos) <= int(cds_length):    
                         csq_record['CDS_RELATIVE_POSITION'] = float(cds_pos/cds_length)

pcgr/oncogenicity.py

@@ -318,21 +318,19 @@ def assign_oncogenicity_evidence(rec = None, oncogenicity_criteria = None, tumor
       ## check if variant has MAF > 0.01 (SBVS1) or > 0.05 in any of five major gnomAD subpopulations (exome set)
       for pop in pcgr_vars.GNOMAD_MAIN_EXON_AF_TAGS:
          if not variant_data[pop] is None:
-            if float(variant_data[pop]) >= pcgr_vars.ONCOGENICITY['gnomAD_very_common_AF']:
+            if float(variant_data[pop]) >= float(pcgr_vars.ONCOGENICITY['gnomAD_very_common_AF']):
                variant_data["ONCG_SBVS1"] = True
-      for pop in pcgr_vars.GNOMAD_MAIN_EXON_AF_TAGS:
-         if not variant_data[pop] is None:
-            if float(variant_data[pop]) >= pcgr_vars.ONCOGENICITY['gnomAD_common_AF'] and variant_data["ONCG_SBVS1"] is False:
+            if float(variant_data[pop]) >= float(pcgr_vars.ONCOGENICITY['gnomAD_common_AF']) and variant_data["ONCG_SBVS1"] is False:
                variant_data["ONCG_SBS1"] = True
-
+               
       approx_zero_pop_freq = 0
       for pop in pcgr_vars.GNOMAD_MAIN_EXON_AF_TAGS:
          ## no MAF recorded in gnomAD for this population
          if variant_data[pop] is None:
             approx_zero_pop_freq = approx_zero_pop_freq + 1
          else:
             ## Extremely low MAF for this population
-            if float(variant_data[pop]) < pcgr_vars.ONCOGENICITY['gnomAD_extremely_rare_AF']:
+            if float(variant_data[pop]) < float(pcgr_vars.ONCOGENICITY['gnomAD_extremely_rare_AF']):
                approx_zero_pop_freq = approx_zero_pop_freq + 1
 
       ## check if variant is missing or with MAF approximately zero in all five major gnomAD subpopulations (exome set)
@@ -384,7 +382,7 @@ def assign_oncogenicity_evidence(rec = None, oncogenicity_criteria = None, tumor
    variant_data['ONCOGENICITY'] = "VUS"
    variant_data["ONCOGENICITY_DOC"] = "."
    variant_data["ONCOGENICITY_CODE"] = "."
-   variant_data["ONCOGENICITY_SCORE"] = "."
+   variant_data["ONCOGENICITY_SCORE"] = 0
    onc_score_pathogenic = 0
    onc_score_benign = 0
 
@@ -460,7 +458,9 @@ def load_oncogenic_variants(oncogenic_variants_fname: str, logger: Logger):
    with gzip.open(oncogenic_variants_fname, mode='rt') as f:
       reader = csv.DictReader(f, delimiter='\t')
       for row in reader:            
-         gene = str(row['entrezgene'])         
+         gene = str(row['entrezgene'])
+         if not 'oncogenic' in str(row['oncogenicity']).lower():
+            continue         
          oncogenic_variants[str(gene) + '-' + str(row['var_id'])] = row
          if not len(row['hgvsp']) == 0:
             oncogenic_variants[str(gene) + '-' + str(row['hgvsp'])] = row

pcgr/pcgr_vars.py

@@ -4,7 +4,7 @@
 
 ## Version - software and bundle
 PCGR_VERSION = __version__
-DB_VERSION = '20250217'
+DB_VERSION = '20250221'
 
 ## Miscellaneous settings
 NCBI_BUILD_MAF = 'GRCh38'
@@ -222,7 +222,7 @@
     r"(stop_(lost|gained)|start_lost|frameshift_|missense_|splice_(donor|acceptor)|protein_altering|inframe_)"
 CSQ_CODING_SILENT_PATTERN2 = \
     r"(stop_(lost|gained)|start_lost|frameshift_|missense_|splice_(donor|acceptor)|protein_altering|inframe_|synonymous|(start|stop)_retained)"
-CSQ_NULL_PATTERN = r"^(stop_gained|frameshift_)"
+CSQ_NULL_PATTERN = r"^(stop_gained|frameshift_)|&stop_gained"
 CSQ_SPLICE_REGION_PATTERN = r"(splice_|intron_variant)"
 CSQ_SPLICE_DONOR_PATTERN = \
     r"(splice_region_variant|splice_donor_variant|splice_donor_region_variant|splice_donor_5th_base_variant)"

pcgrr/NAMESPACE

@@ -8,6 +8,7 @@ export(append_dbmts_var_link)
 export(append_dbnsfp_var_link)
 export(append_drug_var_link)
 export(append_gwas_citation_phenotype)
+export(append_oncogenicity_docs)
 export(append_targeted_drug_annotations)
 export(append_tcga_var_link)
 export(append_tfbs_annotation)

pcgrr/R/data.R

@@ -14,6 +14,10 @@
 "variant_db_url"
 
 
+#' Oncogenicity criteria (ClinGen/CGC/VICC)
+#'
+"oncogenicity_criteria"
+
 #' Fixed data types/categories used for biomarker evidence, e.g. 'types','levels' etc.
 #'
 "biomarker_evidence"

pcgrr/R/input_data.R

@@ -360,6 +360,8 @@ load_somatic_snv_indel <- function(
     callset[['variant_display']] <- callset[['variant']] |>
       pcgrr::append_cancer_gene_evidence(
         ref_data = ref_data) |>
+      pcgrr::append_oncogenicity_docs(
+        ref_data = ref_data) |>
       pcgrr::append_dbmts_var_link() |>
       pcgrr::append_tcga_var_link() |>
       pcgrr::append_annotation_links() |>
@@ -833,7 +835,7 @@ load_dna_variants <- function(
                 c("VARIANT_ID", "ENTREZGENE","BIOMARKER_SOURCE")),
               by = c("VARIANT_ID","BIOMARKER_SOURCE"),
               relationship = "many-to-many") |>
-            dplyr::rename(BIOMARKER_MATCH = BIOMARKER_MATCHTYPE) |>
+            dplyr::rename("BIOMARKER_MATCH" = "BIOMARKER_MATCHTYPE") |>
             dplyr::mutate(BIOMARKER_RESOLUTION = dplyr::case_when(
               stringr::str_detect(.data$BIOMARKER_MATCH,"by_cna_segment") ~ "gene",
               stringr::str_detect(.data$BIOMARKER_MATCH,"by_genomic_coord") ~ "genomic",
@@ -918,26 +920,26 @@ load_dna_variants <- function(
               relationship = "many-to-many"
             ) |>
             dplyr::rename(
-              BM_VARIANT_ID = VARIANT_ID,
-              BM_EVIDENCE_ID = EVIDENCE_ID,
-              BM_SOURCE_DB = BIOMARKER_SOURCE,
-              BM_RESOLUTION = BIOMARKER_RESOLUTION,
-              BM_MATCH = BIOMARKER_MATCH,
-              BM_PRIMARY_SITE = PRIMARY_SITE,
-              BM_EVIDENCE_TYPE = EVIDENCE_TYPE,
-              BM_CANCER_TYPE = CANCER_TYPE,
-              BM_DISEASE_ONTOLOGY_ID = DISEASE_ONTOLOGY_ID,
-              BM_VARIANT_ORIGIN = VARIANT_ORIGIN,
-              BM_EVIDENCE_LEVEL = EVIDENCE_LEVEL,
-              BM_EVIDENCE_DESCRIPTION = EVIDENCE_DESCRIPTION,
-              BM_THERAPEUTIC_CONTEXT = THERAPEUTIC_CONTEXT,
-              BM_CLINICAL_SIGNIFICANCE = CLINICAL_SIGNIFICANCE,
-              BM_CITATION = CITATION,
-              BM_REFERENCE = CITATION_HTML,
-              BM_RATING = RATING,
-              BM_EVIDENCE_DIRECTION = EVIDENCE_DIRECTION,
-              BM_MOLECULAR_PROFILE = MOLECULAR_PROFILE_NAME,
-              BM_MOLECULAR_PROFILE_TYPE = MOLECULAR_PROFILE_TYPE
+              "BM_VARIANT_ID" = "VARIANT_ID",
+              "BM_EVIDENCE_ID" = "EVIDENCE_ID",
+              "BM_SOURCE_DB" = "BIOMARKER_SOURCE",
+              "BM_RESOLUTION" = "BIOMARKER_RESOLUTION",
+              "BM_MATCH" = "BIOMARKER_MATCH",
+              "BM_PRIMARY_SITE" = "PRIMARY_SITE",
+              "BM_EVIDENCE_TYPE" = "EVIDENCE_TYPE",
+              "BM_CANCER_TYPE" = "CANCER_TYPE",
+              "BM_DISEASE_ONTOLOGY_ID" = "DISEASE_ONTOLOGY_ID",
+              "BM_VARIANT_ORIGIN" = "VARIANT_ORIGIN",
+              "BM_EVIDENCE_LEVEL" = "EVIDENCE_LEVEL",
+              "BM_EVIDENCE_DESCRIPTION" = "EVIDENCE_DESCRIPTION",
+              "BM_THERAPEUTIC_CONTEXT" = "THERAPEUTIC_CONTEXT",
+              "BM_CLINICAL_SIGNIFICANCE" = "CLINICAL_SIGNIFICANCE",
+              "BM_CITATION" = "CITATION",
+              "BM_REFERENCE" = "CITATION_HTML",
+              "BM_RATING" = "RATING",
+              "BM_EVIDENCE_DIRECTION" = "EVIDENCE_DIRECTION",
+              "BM_MOLECULAR_PROFILE" = "MOLECULAR_PROFILE_NAME",
+              "BM_MOLECULAR_PROFILE_TYPE" = "MOLECULAR_PROFILE_TYPE"
             ) |>
             dplyr::mutate(
               BM_RATING = dplyr::if_else(

pcgrr/R/mutational_signatures.R

@@ -926,7 +926,7 @@ plot_signature_contributions <- function(
                    ymax = prop_signature_ci_upper),
       width = .3)+
     ggplot2::scale_fill_manual(
-      values = head(
+      values = utils::head(
         pcgrr::color_palette$tier$values,
         NROW(plot_data_per_signature))) +
     ggplot2::theme_classic() +

pcgrr/R/reference_data.R

@@ -426,13 +426,20 @@ load_reference_data <- function(
   for(elem in c('tmb',
                 'mutational_signature',
                 'pathway',
+                'oncogenicity',
                 'hotspot',
                 'protein_domain')) {
 
     fname_misc <- file.path(
       pcgr_db_assembly_dir, "misc", "tsv", elem,
       paste0(elem,".tsv.gz")
     )
+    if(elem == "oncogenicity"){
+      fname_misc <- file.path(
+        pcgr_db_assembly_dir, "misc", "tsv", elem,
+        paste0(elem,".tsv")
+      )
+    }
     check_file_exists(fname_misc)
     pcgr_ref_data[['misc']][[elem]] <- as.data.frame(
       readr::read_tsv(

pcgrr/R/report.R

@@ -726,7 +726,8 @@ load_yaml <- function(yaml_fname, report_mode = "CPSR") {
         TRUE ~ PANEL_NAME
       )) |>
       dplyr::arrange(
-        dplyr::desc(PRIMARY_TARGET), SYMBOL)
+        dplyr::desc(.data$PRIMARY_TARGET),
+        .data$SYMBOL)
 
 
   }

pcgrr/R/variant_annotation.R

@@ -221,6 +221,60 @@ append_gwas_citation_phenotype <-
     }
   }
 
+#' Function that adds oncogenicity documentation from codes
+#'
+#' @param var_df Data frame of sample variants from VCF
+#' @param ref_data PCGR/CPSR reference data object
+#'
+#' @return vcf_data_df
+#'
+#' @export
+append_oncogenicity_docs <- function(
+    var_df,
+    ref_data = NULL){
+
+  if (any(grepl(paste0("^ONCOGENICITY_CODE$"), names(var_df))) &
+      any(grepl(paste0("^VAR_ID$"), names(var_df)))) {
+
+    var_df_unique_slim <- dplyr::select(
+      var_df, c("VAR_ID", "ONCOGENICITY_CODE")) |>
+      dplyr::filter(!is.na(.data$ONCOGENICITY_CODE)) |>
+      tidyr::separate_rows(
+        "ONCOGENICITY_CODE", sep = "\\|") |>
+      dplyr::distinct()
+
+    if (nrow(var_df_unique_slim ) > 0) {
+      var_df_unique_slim  <- var_df_unique_slim  |>
+        dplyr::left_join(
+          dplyr::select(
+            ref_data[['misc']][['oncogenicity']],
+            "CODE",
+            "SCORE",
+            "DESCRIPTION"),
+          by = c("ONCOGENICITY_CODE" = "CODE")) |>
+        dplyr::mutate(
+          ONCOGENICITY_DOC = paste(
+            .data$ONCOGENICITY_CODE,
+            paste0(
+              .data$DESCRIPTION, " (<b>",
+              .data$SCORE,"</b>)"),
+            sep=": ")) |>
+        dplyr::arrange(
+          .data$VAR_ID, dplyr::desc(.data$SCORE)) |>
+        dplyr::group_by(.data$VAR_ID) |>
+        dplyr::summarise(
+          ONCOGENICITY_DOC = paste(
+            unique(.data$ONCOGENICITY_DOC), collapse = ", "))
+
+      var_df <- dplyr::left_join(
+        var_df, var_df_unique_slim,
+        by = c("VAR_ID" = "VAR_ID"))
+    }else{
+      var_df$ONCOGENICITY_DOC <- NA
+    }
+  }
+  return(var_df)
+}
 
 #' Function that adds TCGA annotations (cohort, frequency etc.) to variant identifiers
 #'

pcgrr/data-raw/data-raw.R

@@ -212,6 +212,7 @@ data_coltype_defs[['snv_indel_somatic_raw']] <- readr::cols_only(
   EXONIC_STATUS = readr::col_character(),
   ALTERATION = readr::col_character(),
   PROTEIN_CHANGE = readr::col_character(),
+  GRANTHAM_DISTANCE = readr::col_integer(),
   HGVSp_short = readr::col_character(),
   HGVSc = readr::col_character(),
   HGVSc_RefSeq = readr::col_character(),
@@ -227,6 +228,8 @@ data_coltype_defs[['snv_indel_somatic_raw']] <- readr::cols_only(
   ONCOGENICITY = readr::col_character(),
   ONCOGENICITY_CODE = readr::col_character(),
   ONCOGENICITY_SCORE = readr::col_integer(),
+  KNOWN_ONCOGENIC = readr::col_character(),
+  KNOWN_ONCOGENIC_SITE = readr::col_character(),
   PFAM_DOMAIN = readr::col_character(),
   PFAM_DOMAIN_NAME = readr::col_character(),
   SYMBOL = readr::col_character(),
@@ -314,6 +317,7 @@ data_coltype_defs[['snv_indel_germline_raw']] <- readr::cols_only(
   CODING_STATUS = readr::col_character(),
   EXONIC_STATUS = readr::col_character(),
   PROTEIN_CHANGE = readr::col_character(),
+  GRANTHAM_DISTANCE = readr::col_character(),
   ALTERATION = readr::col_character(),
   HGVSp_short = readr::col_character(),
   HGVSc = readr::col_character(),
@@ -558,6 +562,7 @@ tsv_cols[['snv_indel']] <-
     'ONCOGENICITY',
     'ONCOGENICITY_CODE',
     'ONCOGENICITY_SCORE',
+    #'ONCOGENICITY_DOC',
     'HGVSc',
     'HGVSc_RefSeq',
     'HGVSp',
@@ -782,6 +787,7 @@ dt_display[['snv_indel_gene_actionable']] <-
     'LOF_FILTER',
     'ONCOGENICITY',
     'ONCOGENICITY_CODE',
+    'ONCOGENICITY_DOC',
     'ONCOGENICITY_SCORE',
     'VEP_ALL_CSQ',
     'DBSNP_RSID',
@@ -842,7 +848,7 @@ dt_display[['snv_indel_tier3']] <-
     'PREDICTED_EFFECT',
     'ONCOGENICITY_CODE',
     'ONCOGENICITY_SCORE',
-    'VEP_ALL_CSQ',
+    'ONCOGENICITY_DOC',
     'DBSNP_RSID',
     'CLINVAR',
     'CLINVAR_CLASSIFICATION',
@@ -884,6 +890,7 @@ dt_display[['tier4']] <-
     'REGULATORY_ANNOTATION',
     'ONCOGENICITY_CODE',
     'ONCOGENICITY_SCORE',
+    'ONCOGENICITY_DOC',
     'VEP_ALL_CSQ',
     'DBSNP_RSID',
     'ENSEMBL_GENE_ID',
@@ -1151,3 +1158,9 @@ rm(cancer_phenotypes_regex,
    dt_display,
    tsv_cols,
    c)
+
+oncogenicity_criteria <- readr::read_tsv(
+  "data-raw/oncogenicity.tsv", show_col_types = F
+)
+usethis::use_data(oncogenicity_criteria, overwrite = T)
+

pcgrr/data-raw/oncogenicity.tsv

@@ -0,0 +1,15 @@
+code	score	category	pole	description	resource
+ONCG_OVS1	8	funcvar	P	Null variant - predicted as LoF - in bona fide tumor suppressor gene	VEP;CGC;CancerMine
+ONCG_OS1	4	funcvar	P	Same amino acid change as previously established oncogenic variant - regardless of nucleotide change	ClinVar
+ONCG_OS3	4	funcvar	P	Located in a mutation hotspot with >= 50 samples with variant at AA position, >= 10 samples with same AA change	cancerhotspots.org
+ONCG_OM1	2	funcvar	P	Presumably critical site of functional domain	CIViC
+ONCG_OM2	2	funcvar	P	Protein length changes from in-frame dels/ins in known oncogene/tumor suppressor genes or stop-loss variants in a tumor suppressor gene	VEP;CGC;CancerMine
+ONCG_OM3	2	funcvar	P	Missense variant at an amino acid residue where a different missense variant determined to be oncogenic (using this standard) has been documented	ClinVar
+ONCG_OM4	2	funcvar	P	Located in a mutation hotspot with < 50 samples with variant at AA position, >= 10 samples with same AA change	cancerhotspots.org
+ONCG_OP1	1	funcvar	P	Multiple lines of computational evidence support of a damaging variant effect on the gene or gene product	dbNSFP
+ONCG_OP3	1	funccomp	P	Located in a mutation hotspot with < 10 samples with the same amino acid change	cancerhotspots.org
+ONCG_OP4	1	clinpop	P	Absent from controls (gnomAD) / very low MAF (any five major gnomAD subpopulations)	gnomAD
+ONCG_SBVS1	-8	clinpop	B	Very high MAF (any five major gnomAD subpopulations)	gnomAD
+ONCG_SBS1	-4	clinpop	B	High MAF (any five major gnomAD subpopulations)	gnomAD
+ONCG_SBP1	-1	funccomp	B	Multiple lines of computational evidence support a benign variant effect on the gene or gene product	dbNSFP
+ONCG_SBP2	-1	funcvar	B	Silent and intronic changes outside of the consensus splice site	VEP

pcgrr/inst/templates/pcgr_quarto_report/documentation.qmd

@@ -67,7 +67,7 @@ for(i in 1:NROW(ref_datasets)){
     if(!is.na(license_url)){
       if(source == "cgc" | source == "gepa" | source == "dbnsfp"){
         s <- paste0("    * [", source_full, "](", url, ") - ", description, " (<b>", version, "</b>)",
-                    " - <b>[", license,"](", license_url, ")</b>")
+                    " - [<b>", license,"</b>](", license_url, ")")
       }else{
         s <- paste0("    * [", source_full, "](", url, ") - ", description, " (<b>", version, "</b>)",
                     " - [", license,"](", license_url, ")")