Comparing SNP/indel detection tools on 454

[dePhi]

Hi everybody,

I'm quite new to the field of sequencing and working on a project to find out which software we can use best to detect SNP's and indels using a recently bought 454.
So far I have compiled a small incomplete list of which tools to look at, but these are mostly limited to separate SNP detection tools. I think it is mostly lacking integrated solutions or software packages which also have SNP/indel detection applications besides alignment and/or assembly capabilities.

My list so far:

Seperate tools:

Gigabayes
454SWAP
VAAL

Integrated/package

454 software package
CLCbio Genomics Workbench
SeqMan Ngen

Is it perhaps wise to include alternative base-callers in the comparison as well?

I would be very happy to get any suggestions or tips from people working on 454 data as well.

Thank you for your time and input,

dePhi

[joa_ds]

my experience with the state of the art software is, they all lack something.

Especially the 454 software is not good for resequencing purposes. The AVA software is not able to decently detect gaps/inserts/indels.

Also the mapping software is not very good.

But i understand the manufacturers. Everybody is doing very different experiments, it is difficult to generate a working broad software platform. For general resequencing experiments, the AVA is ok.

For example finding differences in a yeast strain, AVA is ok.

But for analysing complex multiplex amplicon sequencing experiments, AVA is not what you need. In fact, nothing is what you need. You will have to design your own pipeline, design your own database probably if you want to compare different runs with each other etc etc.

The data analysis is not what it should be. We used the sequencer to analyse the PCR efficiency and primer mismatches in the multiplex. I am quite sure no standard package has that kind of software or analyses. But when you design your own databases and use a simple program as BLAT you can get great results to optimize your PCR reactions. We reduced erronouos pcr products from 22% to 4-5% of the total sequence pool in.

Conclusion: You will need a bunch of informatics/servers to analyse your sequencing data properly!

[dePhi]

Cheers mate, thx for the input!

I was already getting the feeling that this isn't going to be the "small project" we thought it would be.
Ah well, here we go!