Skip to content

infer_experiment.py

Infer the strandedness of an RNA-seq experiment from a BAM file. Determines whether the experiment is:

  1. Single-end or paired-end
  2. If strand-specific, how reads were stranded (sense vs antisense)

Usage

infer_experiment.py -i input.bam -r gene_model.bed

Options

Option Description Default
-i, --input-file Alignment file in SAM or BAM format Required
-r, --refgene Reference gene model in BED format Required
-s, --sample-size Number of reads to sample 200000
-q, --mapq Minimum mapping quality for "uniquely mapped" 30

Output

Prints strandedness fractions to stdout:

This is PairEnd Data
Fraction of reads failed to determine: 0.0172
Fraction of reads explained by "1++,1--,2+-,2-+": 0.4903
Fraction of reads explained by "1+-,1-+,2++,2--": 0.4925
  • Values near 0.5/0.5 indicate unstranded
  • Dominant fraction indicates the strand rule to use with other tools