Hi,

I'm using featureCounts (subread v2.0.2) to generate a count matrix for an RNAseq experiment. I have aligned paired-end .sam files sorted by read name and have specified the -p flag and the −−countReadPairs parameter (please see code below). When I run the command, I get ERROR: invalid parameter: '−−countReadPairs'. featureCounts will work just fine if the −−countReadPairs is not included but as I understand it, in version 2.0.2, −−countReadPairs must be included in order to get counts by fragments (as opposed to earlier versions where -p took care of 1) specifying if reads were paired-end & 2) that read pairs were to be counted). A note on the code below: I changed the path names to the .gtf, output .txt and sorted.sam files in this post for privacy reasons.

Thank you so much for your help! Sandra

$ featureCounts -T 4 -s 2 -p \
--countReadPairs \
-a /path/to/genes.gtf \
-t exon \
-g gene_id \
-o /path/to/featurecountsPE.txt /path/to/aligned/*.sorted.sam \
2> /path/to/aligned/results/counts/featurecountsPE.screen-output1.log

Hi Sandra,

I tested the source code version, the macOS version and Linux version of subread-2.0.2 using the same command as yours, and I didn't get this error message. The --countReadPairs option worked as expected.

Your command line seems correct.

Can you try this command:

featureCounts -v

to see the version of featureCounts?

Also, "−" and "-" are different symbols, and I saw you used the unicode "−" symbol in your post. Please be aware that some PDF files replace the ASCII symbol of "-" with the unicode symbol of "−". When you copy some options from the PDF file to your command, there is a chance that you actually copied the unicode "−" symbol, which isn't recognised by featureCounts as the prefix of options. So please make sure your command included the ASCII "-" symbol instead of the unicode "−" symbol as the prefix of the --countReadPairs option.

Cheers, Yang