HTS Mappers!

[dan]

Anyone vounteer to update the wiki ;-)

http://wwwdev.ebi.ac.uk/fg/hts_mappers/

http://dx.doi.org/10.1093/bioinformatics/bts605

Motivation: A ubiquitous and fundamental step in high-throughput sequencing analysis is the alignment (mapping) of the generated reads to a reference sequence. To accomplish this task numerous software tools have been proposed. Determining the mappers that are most suitable for a specific application is not trivial.

Results: This survey focuses on classifying mappers through a wide number of characteristics. The goal is to allow practitioners to compare the mappers more easily and find those that are most suitable for their specific problem.

Availability: A regularly updated compendium of mappers can be found at http://wwwdev.ebi.ac.uk/fg/hts_mappe...g/hts_mappers/.

Contact: [email protected]

[adaptivegenome]

A comparison of accuracy would have been nice...

[dan]

I was thinking to organize a 'mapathon' on the wiki.... how would you recommend benchmarking accuracy?

[adaptivegenome]

I was thinking to organize a 'mapathon' on the wiki.... how would you recommend benchmarking accuracy?

That would be AWESOME! I think we should query SEQanswers folks and vote on a number of metrics... this would be fun group project...

[dan]

Cool! It shouldn't be too hard either, but we need to have clear guidelines about what to measure and how, so that results are comparable across mappers. The only other thing we need is a dataset and a place to host it...

[bacdirector]

I'd recommend using % reads mapped and the # of reads mapped uniquely as part of the evaluation, and keep in mind that the # of known heterozygotes is a measure of sensitivity, NOT accuracy. My two cents.

[dan]

Should we generate data so we have a 'gold standard'? Or just work with metrics based on consistency, speed, and coverage?

[adaptivegenome]

Are you talking about simulated or real?

[dan]

Simulated reads from real genomes.

We could simulate a range of 'noise' across a series of error models and produce reads characteristic of a variety of technologies... there must be software to do this already right?

[adaptivegenome]

There is and we have these datasets already. I would be happy to share. We did a number of tests with this dataset for a manuscript we are sending out.

[dan]

Fabulous! Please link to the datasets, and I'll try to put wheels in motion via the wiki :-)