Mapping very short (<20 bp) reads

[hclee]

Hi,

I've been a SOAP2 user but recently got some very short reads (<20 bp) to map to a reference genome. Can someone please suggest an aligner which can handle the problem? SOAP seems to map only reads >30 bp.

Thank you.

[Kennels]

Hi,

I've been a SOAP2 user but recently got some very short reads (<20 bp) to map to a reference genome. Can someone please suggest an aligner which can handle the problem? SOAP seems to map only reads >30 bp.

Thank you.

You could try bowtie or bwa

[sdarko]

Command-line BLAST could work well.

[husamia]

I am curious about your data and brings unrelated question? Aren't <20bp reads considered ambigous? Technically its possible that good amount of your reads align to multiple regions in genome. Hence (ambiguous). Is there data to support what is a good enough read length that provides least amount of ambiguous-ness?
I was playing around with searching and as long as I enter random sequence shorter than 20bps in general I get multiple perfect match with genome.

[mnkyboy]

I am curious about your data and brings unrelated question? Aren't <20bp reads considered ambigous? Technically its possible that good amount of your reads align to multiple regions in genome. Hence (ambiguous). Is there data to support what is a good enough read length that provides least amount of ambiguous-ness?
I was playing around with searching and as long as I enter random sequence shorter than 20bps in general I get multiple perfect match with genome.

I was wondering about this too. Shorter than 20 you cannot differentiate them from random sequences.