Reads longer than 75bp

[cormicp]

Hi all,

Am relatively new to NGS and our group has recently moved to using 80bp reads (not paired-end reads). What I am wondering simply is what programs people are using to align reads of these lengths to a reference genome. To my naive eye eland does not seem to be ideal for reads greater than 40bp in length. Probably should say we are generating 80bp reads from both DNA and RNA seq experiments so alignment advice for both would be appreciated.

Thanks in advance for any advice given.

[nilshomer]

Hi all,

Am relatively new to NGS and our group has recently moved to using 80bp reads (not paired-end reads). What I am wondering simply is what programs people are using to align reads of these lengths to a reference genome. To my naive eye eland does not seem to be ideal for reads greater than 40bp in length. Probably should say we are generating 80bp reads from both DNA and RNA seq experiments so alignment advice for both would be appreciated.

Thanks in advance for any advice given.

There are many whole-genome alignment algorithms that have come out in the past year and a half. I am biased since I have my own BFAST. Others include but are not limited to Bowtie, BWA, MAQ, SHRiMP, SOAP. For a more complete list see:
http://seqanswers.com/forums/showthread.php?t=43

They all perform fairly well (based on my simulations), with some using less memory at the cost of speed, some faster at the cost of memory, some find indels (some do not), some are better with ABI SOLiD data, and so on.

Let the debate begin in your own mind.

[der_eiskern]

There are many whole-genome alignment algorithms that have come out in the past year and a half.
They all perform fairly well (based on my simulations), with some using less memory at the cost of speed, some faster at the cost of memory, some find indels (some do not), some are better with ABI SOLiD data, and so on.

Let the debate begin in your own mind.

which ones work best do you think with SOLiD data?

[nilshomer]

which ones work best do you think with SOLiD data?

This is totally biased, and I mean completely biased. I am the author of BFAST, the BLAT-like fast accurate search tool, which I believe it is the best SOLiD (and Illumina) alignment tool. For SOLiD, it has the proper local alignment (see this paper, which I am an author) which considers indels, SNPs, and color errors simultaneously (analogous color space smith waterman). Next, it is flexible, allowing it to be tuned for any sensitivity (by creating multiple indexes of the reference). It also scales well with read length, and is above all fast and accurate.

To give it a test run, I have a demonstration server called the BFAST web server. PM me if you have any more questions.