Description

Intron predictions from short-read RNA-Seq experiments

Methods

• The reads from 432 single-ended and paired-end Poly(A)_ short-read RNA-Seq experiments were obtained from ENCODE. The selection emphasized brain, testis, liver, or the H1 cell line to optimize transcript diversity, as well as reads for at least 100bp. The experiments included are lists below.
• Reads were aligned to the T2T CHM13 assembly using STAR 2.7.5. with parameters derived from the ENCODE pipeline. The maximum number of allowed multi-mapping reads was increases to 2048 to improve sensitivity over specificity. The intention is to develop post-mapping filters to obtain more specific mappings.

  --outFilterMultimapNmax 2048	--alignSJoverhangMin 8	--alignSJDBoverhangMin 1
  --outFilterMismatchNmax 999	--outFilterMismatchNoverReadLmax 0.04	--alignIntronMin 20
  --alignIntronMax 1000000	--alignMatesGapMax 1000000	--outSAMunmapped Within
  --outFilterType BySJout	--outSAMstrandField intronMotif (single ended)	--sjdbScore 1