Entering edit mode
4.5 years ago
bertb ▴ 10
I am trying to run a macs2 peak analysis on bam files generated from unpaired reads. Format = AUTO, with no additional options invoked. The warning I receive is "Too few paired peaks (0) so I can not build the model! Broader your MFOLD range parameter may erase this error..."
Re running the script with --nomodel and a suggested --extsize provides strange data that does not line up with what I can see on IGV.
I suspect the issue is that macs is trying to build a model with paired reads when I haven't provided paired data - is this possible? Should I be adjusting my usage parameters?
This has nothing to do with paired-end sequencing. The paired peaks are based on the shifting model that MACS builds based on tags aligned to the plus and minus strand. Please read the MACS paper. Can you post a screenshot of the genome browser? Maybe the data simply do not contain peaks (be it for a biological or technical reason). What was the target of the ChIP (which protein)?
Thanks for the fast response - I'll try and get a screenshot up. Appreciate the help. (Originally posted as answer in error).
I'm not able to post a screenshot, but when I load the bam and bedgraph files in the genome browser, the bedgraph files basically look like a compressed version of the bam files. The background is higher, and peaks that appear more than 5-fold above background in the bam files are less than 2-fold enriched in the bedgraph files.
Thanks in advance for any suggestions to help generate a more accurate peakset.
See this: How to add images to a Biostars post