GENCODE.v38: # Import GENCODE v38 annotation. GENCODE.v38.GTF <-...

In J0bbie/R2CPCT: Misc. functions to perform import and analysis of WGS and RNA-Seq samples

Description

# Remove pseudogenes and assorted non-interesting types. tmp <- tmp[!grepl('pseudogene|nonsense_mediated_decay|TEC', tmp$gene_type) | tmp$gene_type == 'polymorphic_pseudogene',]

Usage

data(GENCODE.v38)

Format

GRanges object containing the cleaned-up genes from GENCODE v38.

Details

# Remove clone-based genes. tmp <- tmp[!grepl('^AC[0-9]|^AP00|^RP11-', tmp$gene_name)]

# Determine number of unique exons per gene. exonInfo <- tibble::as_tibble(S4Vectors::mcols(tmp[,c('exon_id', 'gene_id', 'transcript_type')])) exonInfo <- exonInfo dplyr::group_by(gene_id) dplyr::filter(!grepl('pseudogene|nonsense_mediated_decay|TEC|retained_intron|non_stop_decay', transcript_type) | transcript_type == 'polymorphic_pseudogene') dplyr::summarize(totalExonsInGene = dplyr::n_distinct(exon_id, na.rm = TRUE))

tmp$totalExonsInGene <- exonInfo[base::match(tmp$gene_id, exonInfo$gene_id),]$totalExonsInGene

# Only retain the correctly-placed genetic elements. tmp <- tmp[tmp$level

# Only retain genes with at least one non-suspect transcript or are single exons. tmp <- tmp[tmp$totalExonsInGene >= 1,]

# Only retain the gene-level elements. tmp <- tmp[tmp$type == 'gene',]

# Remove unused columns. S4Vectors::mcols(tmp) <- S4Vectors::mcols(tmp)[!grepl('remap_|havana_|phase|score|ont|hgnc_id|ccdsid|transcript|exon|protein|gene_status', colnames(S4Vectors::mcols(tmp)))]

# Convert columns. tmp$source <- as.character(tmp$source) tmp$type <- as.character(tmp$type) tmp$ENSEMBL <- as.character(tmp$gene_id); tmp$gene_id <- NULL tmp$SYMBOL <- as.character(tmp$gene_name); tmp$gene_name <- NULL

# Convert to factor. S4Vectors::mcols(tmp) <- apply(S4Vectors::mcols(tmp), 2, function(x) factor(as.character(x))) S4Vectors::mcols(tmp) <- droplevels(S4Vectors::mcols(tmp)) GenomeInfoDb::seqlevels(tmp) <- base::paste0('chr', GenomeInfoDb::seqlevels(tmp)) tmp$ENSEMBL <- as.character(tmp$ENSEMBL)

# Retrieve the common gene name. tmp$SYMBOL <- as.character(tmp$SYMBOL) commonSYMBOL <- limma::alias2SymbolTable(tmp$SYMBOL) tmp$SYMBOL <- base::ifelse(is.na(commonSYMBOL), tmp$SYMBOL, commonSYMBOL)

# Save to package. GENCODE.v38 <- tmp

usethis::use_data(GENCODE.v38, overwrite = TRUE)

Source

http://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_38/GRCh37_mapping/gencode.v38lift37.annotation.gtf.gz

J0bbie/R2CPCT documentation built on Feb. 24, 2022, 8:15 a.m.