Short sequence alignment

[rboettcher]

Hi all,
I am currently looking for a software to do alignments of short reads against other short sequences (so not the whole genome). The only problem is: I want to do a lot of these alignments. So instead of aligning Xmio reads to one big sequence, my goal is to align them against 10k (or so) very short sequences. Does anybody know a suitable software?

Regards

€: I should add that I'm not aiming on multiple sequence alignments.

[henry.wood]

We have done something like this using bwa. We made an artificial reference 'genome' from 3000 short (5-15Kb) fragments treating each fragment as a chromosome. We then aligned several lanes of Illumina data against this quite happily.

[rboettcher]

I would prefer something not relying on prior indexing, so a fast implementation of Needleman-Wunsch or Hirschberg-algorithm would be sufficient I suppose, as long as it is accessible by console.

[zee]

Novoalign does full Needleman-Wunsch after seeding matches. You can also map short reads to small fragments.

[sparks]

Adding to Zees comment, Novoalign indexing step is very fast for small references so it shouldn't be an issue.

I was also wondering if you're interested in alignments that overlap rather than are contained in your short target sequences. If you are then you could add N's to both ends of the short sequences that you index.

[rboettcher]

No overlaps, just aligning the sequences to a number of short reference sequences. Unfortunately, I couldn't get Novoalign to run on my system

[sparks]

What problem did you have with Novoalign?

[ericgdp]

May be you can use Abacas.

[biznatch]

PatternHunter?

http://www.bioinformaticssolutions.com/products/ph/

[rboettcher]

Thank you all very much for your suggestions, but I guess none of the aligners mentioned is really meant to do what I intend it to do, as far as I can tell from studying the manuals/faqs.
So after some consideration, I decided to implement a simple NW-algorithm and after switching from Perl to Java the speed is acceptable.
I had hoped I could skip this step but it seems there is no real alternative.
Thanks anyway, I guess you will read about my next problems soon enough on the forums

Best Regards