search

katholt / RedDog

34 stars 4 forks source link

Could I use RedDog to map sequence reads to few genes (not the whole genome)? #65

Open MostafaYA opened 5 years ago

MostafaYA commented 5 years ago

Hi, I have the sequence reads of some strains (illumina reads). I want to map these sequences to certain genes (not the whole genome of the reference) and to determine if there are SNPs introduced and whether these SNPs result in aminoacid change or not.

Could I tweak the parameters of RedDog so I can use it to map sequence reads to certain genes? As another option to me is srst2, however it is not giving detailed information about SNPs. Ariba also sometimes for reasons I don't know does not detect SNPs which I can easily see in the assembled genomes.

What could be the other alternative for mapping strain reads against genes?

Thanks and regards,

d-j-e commented 5 years ago

You could concatenate the genes together into a 'short' GenBank reference - maybe adding twenty to fifty bases up and downstream to each. Reddog uses the 'local sensitive' setting for Bowtie so should handle this situation.

Do keep in mind that whilst reddog doesn't examine the indels, it does keep the ones that pass the filtering steps, so you may want to check to vcfs for any potential indels as well...

MostafaYA commented 5 years ago

@d-j-e thanks a lot for your answer. It worked for me with some tweaking. However, the target SNP I was looking for was not reported by RedDog. May be that the reddog filteration criteria are very strict