Adjusting for uneven Nextseq loading

Question

[rbronste]

Ran a ChIP-seq experiment with cells treated with vehicle and a drug, so sequenced 4 samples (vehicle input, vehicleIP, drugInput, drugIP). During multiplexing the vehicle input was unevenly added to the pooled libraries, and following sequencing therefore had a much bigger fastq file. I am wondering about usability of these results and how to potentially adjust during analysis for this uneven loading. Thank you.

p.s. The 3 smaller fastq files still gave a good amount of aligned reads so thats not an issue. Just thinking in terms of peak calling, differential peak analysis etc.

Answer 1

[colindaven]

Just downsampling the big fastq might be a good plan. You can use samtools for this. I'd visualize all the results too before analysis.

I find Chip-seq to be notoriously dodgy.

[rbronste]

So you would just subsample at different % until it hits around the size of the other BAM files or is there another downstream metric?

[colindaven]

Yep, sounds about right!