This is a test version of Biostars. For the public version, visit https://www.biostars.org.
count reads in ambiguous regions for both features

Hi,

According to the last example of this figure reads that fall in overlapping regions of 2 features are discarded in every mode of htseq-count. I understand how this is reasonable in DE analysis. However, I am now interested in raw counts of features in my experimental setup as they are just predictions and I want to count a read that falls in an ambiguous region for both features. Is there any other tool or R package out there that allows this?

Thanks

rna-seq read-count
• 6,025 views • link
updated by Ram Ram • written by caner • caner

2 answers

Quick Google search pointed me to this tool: http://bioinf.wehi.edu.au/featureCounts/. It is said on their page: "By default, featureCounts does not count reads overlapping with more than one feature (or more than one meta-feature when summarizing at meta-feature level). Users can use the -O option to instruct featureCounts to count such reads (they will be assigned to all their overlapping features or meta-features)." To my knowledge, other tools such as htseq-count or summarizeOverlaps in R will only count one feature even in their less strict modes.

• 0 views • link
updated by Ram Ram • written by mkulecka mkulecka

+1 for featureCounts, it works well and is well documented.

• 0 views • link
updated by Ram Ram • written by Carlo Yague Carlo Yague

Hi guys,

I'm using featureCounts with default parameters in my samples (paired-end reads) and I got around 30% of Successfully assigned fragments, but when I use the option -O (overlapping reads) the percentage increase up to 80%.

I'm working with Human genome and I would like your opinion about the use or not the option -O. Some tools has the option to choose one of the mapping region and not counting the other ones, but it looks like that featureCounts doesn't have this option.

Any hints?

Thanks,
Milton

• 0 views • link
updated by Ram Ram • written by yutakajr • yutakajr

I think you should ask a separate question.

Anyway, my opinion is that the use of -O depends on the annotation you use. In other words, on what features will you count reads ? Are you looking only at protein coding genes expression? Or also on non-coding RNA genes, XUT, CUTS, ... ? In the first case you should expect few overlap and probably shouldn't use the -O option.

• 0 views • link
updated by Ram Ram • written by Carlo Yague Carlo Yague

Log in to answer this question.

More posts like this