Question

using only uniquely-mapping reads for alternative splicing using SGSeq

0

Entering edit mode

116100016 • 0

@116100016-22156

Last seen 5.0 years ago

I am using SGSeq for alternative splicing analysis of the androgen receptor and was wondering if there was a way to tell the program only to use uniquely mapped reads from the BAM file in its analyses. As there seems to be discrepancy with IGV counting which I'd like to try to tease out.

rna seq SGSeq Alternative splicing • 966 views

ADD COMMENT • link updated 5.1 years ago by Leonard Goldstein ▴ 260 • written 5.1 years ago by 116100016 • 0

score 0 · Answer 1 · 2019-10-16

0

Entering edit mode

Leonard Goldstein ▴ 260

@leonard-goldstein-6845

Last seen 6 months ago

Australia

You can influence which reads are analyzed when setting up the data frame with sample information. To do this column file_bam needs to be a BamFileList and sample information must be stored in a DataFrame object. Please see the manual page for Rsamtools::BamFileList for more details.

ADD COMMENT • link 5.1 years ago Leonard Goldstein ▴ 260

0

Entering edit mode

Actually looking at the code in more detail, BamFile details from the sample information currently do not propagate. The hard-coded behavior is that only primary alignments in the BAM file are considered for analysis. In other words, each read should be analyzed only once, but the analysis may include reads that do not align uniquely. For your question it may be easiest to pre-filter BAMs to only contain uniquely mapped reads.

ADD REPLY • link 5.1 years ago Leonard Goldstein ▴ 260