kostkalab / aenmd_data

aenmd_data

Data packages for the aenmd R package

• Introduction
• Installation
• Example
• Additional information

Introduction

This repository contains R data packages for the aenmd package; aenmd annotates transcripts with variants that cause premature termination codons with regard to predicted escape nonsense-mediated decay (NMD). aenmd uses data objects derived from transcript models, and here we provide them for transcript sets for the GRCh37 and GRCh38 assemblies. This makes it straight-forward to use aenmd with variant annotations from either assembly.

Currently, the repository contains three data packages:

• aenmd.data.ensdb.v105 High-confidence transcript set based on Ensembl version 105. High-confidence means protein-coding transcripts with transcript support level 1 (or single exon transcripts).

• aenmd.data.gencode.v43 Protein-coding transcripts from GENCODE version 43.

• aenmd.data.gencode.v43.grch37 Protein-coding transcripts from GENCODE version 43, mapped to GRCh37.

Installation

Install packages by selecting the appropriate subdirectory from this aenmd_data repository.

#- need the remotes package
if(!require(remotes)){
    install.packages("remotes")
}

#- install ENSEMBL v105 transcripts on GRCh38
remotes::install_github(repo = "kostkalab/aenmd_data",
                          subdir = "aenmd.data.ensdb.v105")

#- install GENCODE v43 transcripts on GRCh38
remotes::install_github(repo = "kostkalab/aenmd_data",
                          subdir = 'aenmd.data.gencode.v43')

#- install GENCODE v43 transcripts on GRCh37
remotes::install_github(repo = "kostkalab/aenmd_data",
                          subdir = 'aenmd.data.gencode.v43.grch37')

Annotating variants from GRCh37 or GRCh38 assemblies

library(aenmd)

#- load variants (two assemblies)
vars_grch38 <- system.file('extdata/clinvar_20221211_noinfo_sample1k.vcf.gz', package = 'aenmd') |> 
               aenmd:::parse_vcf_vcfR()

vars_grch37 <- system.file('extdata/clinvar_grch37_20221211_noinfo_sample1k.vcf.gz', package = 'aenmd') |> 
               aenmd:::parse_vcf_vcfR()

#- ANALYZE GRCH38 VARIANTS
#-------------------------

#- per default we have
ad_get_packagename()
#[1] "aenmd.data.ensdb.v105"
ad_get_genome()
#[1] "GRCh38"

vars_rng_38 <- process_variants(vars_grch38$vcf_rng)
res_38      <- annotate_nmd(vars_rng_38)

#- ANALYZE GRCH37 VARIANTS
#-------------------------

#- change to a GRCh37 annotation package
ad_swap_annotation('aenmd.data.gencode.v43.grch37')

ad_get_packagename()
#[[1] "aenmd.data.gencode.v43.grch37"
ad_get_genome()
#[1] "GRCh37"

#- process variants
vars_rng_37 <- process_variants(vars_grch37$vcf_rng)

#- filter out out variants where we don't know the alternative 100%
n_N         <- Biostrings::alphabetFrequency(vars_rng_37$alt) [,-(1:4)] |> rowSums()
ind_out     <- which(n_N > 0)
vars_rng_37 <- vars_rng_37[-ind_out]

#- finally, annotate on GRCH37
res_37  <- annotate_nmd(vars_rng_37)

Additional information

Each annotation package provide transcript information, which in turn is available via the aenmd package to the user. Typically the functions below are used by aenmd internally, but users can use them to get information about aenmds transcript set, and about PTC-generating SNVs. Below, transcripts are identified by their GENCODE/ENSEMBL transcript ID, e.g. txname <- ENSG00000187634.13; SNVs are identified by their location and alleles, e.g. snvname <- 10:000047074|C|A.

Information about transcripts:

• Transcripts in transcript set. ad_get_txs() List of transcripts in transcript set.
• Transcript coding exons. ad_get_exns_by_tx(txname) Coding exons for a ranscript.
• Transcript coding sequence. ad_get_cds_by_tx(txname) Coding (reference) sequence for a transcript.
• Transcript mask. ad_get_txs_mask() Reduced genomic ranges for coding sequence considered in that package.
• Single exon transcripts. ad_is_single_exn_tx(txname) Query whether a transcript only has a single exon.
• Splice regions. ad_get_splice_mask() Reduced genomic ranges for for splice regions.

Information about PTC-generating SNVs:

• PTC-generating SNVs. ad_is_ptc_snv(snvname) Does a SNV generate a PTC in a transcript?