Variant representation in pooled PCR products

Question

[cjmcdermottroe]

Hi all. I have a plasmid into which I cloned a chemically-synthesized oligo pool. The oligos differ from one another by single nucleotides and hence the resulting plasmid is highly complex/diverse in this particular region (which just so happens to be a gene of interest). I PCR'd across this hypervariable segment and deep-sequenced the amplicon and am now scratching my head as to how I should analyze the data. Does anyone have any software suggestions? I guess I was thinking there might be a way to automate a grep command so it systematically searches the fastq file for the informative segment of each oligo sequence. Can provide more info if required. Thanks!!!

[Pierre Lindenbaum]

and am now scratching my head as to how I should analyze the data

what kind of analysis ?

Answer 1

[ATpoint]

You could make a multifasta with every oligo sequence that you expect to be there. Then build an index from it with e.g. bowtie and align your reads on it. Choose parmeters as strict as possible, so gap opening and mismatch penalty to 1000 or so. This will ensure perfect matches between the reads and the oligo reads. As the oligos are probably short(er) that the read length, you may want to trim your reads to be free of any adapter sequences. I prefer skewer for trimming, but any tool will do. After alignment, index the bam file and use samtools idxstats. It will tell you how many reads have aligned to every reference regions. I guess you only want to make sure that every oligo is present?

[cjmcdermottroe]

Awesome ATPoint, a custom index and alignment is far more sensible than grep. I'll give it a go and report back. Thanks..