Tools for DNA mapping

[remimaglione]

Hi everybody,

I'm new in the NGS. I'm working on DNA mapping with data coming from shotgun sequencing (with illumina technology).
I'm already working with BWA tools for the mapping and if understand good, this one is designed for deal with some big reference genome (like the human one?!).
But I'm looking for program designed to work with small genomes, such as circular genome between 100 and 500kb.

Can anyone help me?

In advance thank you

[chadn737]

All the programs, including BWA, will handle small genomes with ease.

[remimaglione]

Thank you,

but what do you mean by "all the program" ?

Now i'm trying to work with Razers3, but anybody can give me more tools ? ( free tools is prefer )
Someone working with small genome would use a specific program?

[chadn737]

Small reference size is not really an issue. Large references are problematic because your search space is so vast. You don't need a special alignment tool for a small reference Any of the commonly used ones will work on a small references. The two most commonly used are probably BWA or Bowtie/Bowtie2.

[shi]

Thank you,

but what do you mean by "all the program" ?

Now i'm trying to work with Razers3, but anybody can give me more tools ? ( free tools is prefer )
Someone working with small genome would use a specific program?

You may try Subread - http://subread.sourceforge.net

Cheers,
Wei

[remimaglione]

Thank you for this. I will try to do a little simulation on my laptop tonight.
other programs are welcome !

[mastal]

Tools for DNA mapping

See the wiki for a list of available software.

http://seqanswers.com/wiki/SEQanswers

[oiiio]

and see GCAT for benchmarks of different tools

http://www.bioplanet.com/gcat

[remimaglione]

See the wiki for a list of available software.

http://seqanswers.com/wiki/SEQanswers

I have already see this, but it was difficult for me to target a specific free tools for my study.
This is why I created this post to have the opinion of the community

[jgibbons1]

You may want to check out Heng Li's (BWA creator) ROC curve benchmark on read mapping software:

http://lh3lh3.users.sourceforge.net/alnROC.shtml

One of the benefits of BWA, Bowtie2 and SOAP2 is that they use a Burrows Wheeler Transform to index the genome...this is very memory efficient for large genomes (humans for example). Because you're working with a relatively small reference genome, mappers which has-index the genome will work well as well, without a huge memory requirement (MAQ, Cloudburst, ZOOM, Stampy etc.).

I'm not very familiar with the merge-sorting method, but you may want to look into Slider .

I guess a better question is to ask what are you trying to get out of the mapping? Some software is better for SNP discovery, others are better for indels and CNVs and you'll see a trade-off between false-positives rate and false-negative rates.

Here's a nice paper that may help you make your decision:

http://bioinformatics.oxfordjournals.org/content/27/20/2790.short

Cheers.

[jgibbons1]

I should say that I typically use Bowtie2 and BWA.

[oiiio]

which aligners trade off between SNP discovery and indels/CNVs ?? Except in some types of indels, this is certainly a problem at the level of entire pipeline..

[jgibbons1]

SOAP2 for example has great accuracy in calling SNPs but can not confidently call indels larger than 6bp (see link to the Ruffalo et al. 2011 paper above). Novoalign, BWA and Bowtie are better at calling indels (albeit far from perfect).

[oiiio]

I've not tried SOAP2 actually, but I have seen a bunch of comparisons of variant calling pipelines using the other ones you mentioned:

http://www.bioplanet.com/gcat/reports/101/variant-calls/illumina-100bp-pe-exome-30x/bowtie-gatk_ug

Anyways, it all goes to show that you need to tailor your pipeline to your sample for optimal results

[jgibbons1]

Anyways, it all goes to show that you need to tailor your pipeline to your sample for optimal results

I couldn't agree more...