What tool can align a ton of NGS data from a large region to a tiny reference?

[Gingeneticist]

Hi all,

I have NGS data from a patient: Genomic DNA fragments ranging from 300-600bp sequenced on the MiSeq (2x250bp reads) prepared with SureSelect sample prep/target enrichment.

The DNA was enriched for a 4Mb region, but I have focussed in on a smaller anomaly and need a way of pulling out data from one small and specific area. Basically, I want to line up all the reads that contain a certain 37bp sequence and reject everything else.
So far I have tried two bits of software:
Nextgene - great for big alignments but not good for this task as I can't specify this small a region to align my data to
Sequencher - this should be able to do it in theory but can't seem to handle the amount of data I'm asking it to align and crashes


Would greatly appreciate all help and suggestions for what software I could try.

Best wishes,
Claire

[boetsie]

If you just want an exact match of the 37bp, simply use 'grep'.

If mismatches should be allowed, I think you can do this with bowtie2. Make an index of your 37bp sequence and align the paired-reads with the --local command. Be careful though, the --local can also align smaller regions. You should probably filter those alignments.

Regards,
Boetsie

[Gingeneticist]

Thanks boetsie!

[GenoMax]

Using "grep" (to identify the reads containing the exact sequence of interest) followed by alignment with Sequencher should work (at least on subsets of reads if there are millions).

You may want to do some kind of de-duplication (post "grep" selection and) prior to alignment (unless you are interested in counts) to simplify downstream processing/display.

[JamieHeather]

For something like this, you can also use agrep to allow mismatches.