evalModel: Evaluate transcript model

In pram: Pooling RNA-seq datasets for assembling transcript models

Description

Evaluate trascript model's precision and recall on exon nucleotides, splice junctions, and splice patterns by comparing them to transcript targets

Usage

evalModel(model_exons, target_exons)

## S4 method for signature 'GRanges,GRanges'
evalModel(model_exons, target_exons)

## S4 method for signature 'character,character'
evalModel(model_exons, target_exons)

## S4 method for signature 'data.table,data.table'
evalModel(model_exons, target_exons)

## S4 method for signature 'character,data.table'
evalModel(model_exons, target_exons)

Arguments

model_exons	genomic coordinates for transcript model exons
target_exons	genomic coordinates for transcript target exons

Value

a data table of precision, recall, number of true positive, false negative, false positive for all three evaluated features

Methods (by class)

• model_exons = GRanges,target_exons = GRanges: Both model_exons and target_exons are GRanges objects to define genomic coordinates of exons. Required to have a meta-data column named 'trid' to define each exon's transcript ID.

• model_exons = character,target_exons = character: Both model_exons and target_exons are GTF files with full names. Each GTF file is required to have a 'transcript_id' tag in column 9.

• model_exons = data.table,target_exons = data.table: Both model_exons and target_exons are data.table objects to define exon genomic coordinatess. Required to have the following columns:

  • chrom: exon's chromosome, e.g. 'chr8'

  • start: exon's start position

  • end: exon's end position

  • strand: exon's strand, '+' or '-'

  • trid: exon's transcript ID

• model_exons = character,target_exons = data.table: The model_exons is a GTF file with full name and target_exons is a data.table object. Requirements for GTF and data.table are the same as above

Examples

fmdl = system.file('extdata/benchmark/plcf.tsv', package='pram')
ftgt = system.file('extdata/benchmark/tgt.tsv',  package='pram')

mdldt = data.table::fread(fmdl, header=TRUE, sep="\t")
tgtdt = data.table::fread(ftgt, header=TRUE, sep="\t")


evalModel(mdldt, tgtdt)

pram documentation built on Nov. 8, 2020, 11:08 p.m.