What Is The Recommended Way To Use Bwa On Files Containing Haplotypes Of An Exome

Question

[agatorano]

Hello

I am unsure how best to use BWA on two files containing haplotypes of an exome. There are three BWA types. I assume I am to be using the MEM algorithm but I am unsure what many of my options should be. For example what should my gap costs be? match? Are these things well established? what are rules to use when trying to decide on the parameters for this algorithm?

Thank you for your help.

Answer 1

[Leandro Lima]

Hello!

Maybe this post can help you.

What is the best pipeline for human whole exome sequencing?

[agatorano]

This is really helpful in the long term but even they admit to not knowing the parameters to use and why.

[tommivat]

Also, this post is from time before BWA mem algoritm so it does not answer to the question of OP in that way.

Answer 2

[onter]

Typically the default parameters have been tested and optimized on the data they are prescribed to. For example, BWA-MEM is excellent for 75+bp reads. There are also tools online for determining which kind pipeline is suitable for your sample - one of these is GCAT on bioplanet.com which allows you to view benchmarks on pipelines as well as run your own. Below is a GCAT summary of a BWA pipeline comparison with GATK, Freebayes, and ATLAS on a 30x illumina exome:

View Full GCAT Variant Calls Report: Bwa+Atlas (illumina-100bp-pe-exome-30x)

If you are uncertain about what pipeline or parameters to use, I really suggest you give GCAT a shot

[agatorano]

This is extremely interesting