Entering edit mode
8.2 years ago
Jon17
▴
20
I'm trying to setup a Bismark pipeline for a sequence capture project. Does anyone know of example data and results so I can see if I'm getting the right results on full sized data using industry standard analysis of methylation capture data?
Either of these will work
- Roche Nimblegen CpGiant
- Agilent SureSelect MethylSeq
You're unlikely to find a before/after repository to test your Bismark is running correctly, however i'd recommend you take a look at bwameth over Bismark. I (personally) didn't have a good experience with Bismark.
What issues did you run into? Why is bwa-meth >> bismark?
For me, the reasons was that bismark was incredibly slow, and the BAM files it returned where failing picard's ValidateSam. The slowness of Bismark cannot be underestimated.
If you are going to be mapping a lot of bisulphite sequencing then you'll actually want to take a look at bison (http://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-15-337), as bison and bwa-meth seem to be similar in terms of output quality, however bison is designed to run over a cluster (if you have one).
I chose to use bwa-meth (https://github.com/brentp/bwa-meth) as the installation was simple and I was up and processing WGBS in literally minutes after first visiting the github page. I later used pileOmeth and biscuit just as the github page prescribes, and everything worked out just great.