Anyone out there still working with SSR loci? Frustrated with allele size migration I have started amplicon sequencing my loci instead. I guess I hoped to distinguish more alleles, but now I am getting SNPs in flanking sequences that mean few alleles are shared among populations. Some of these look convincing, but I think some of this is noise. Are there substutions to expect as a result of the chemistry? I will probably have to replicate my data, but any tweaks?
Solution would be very dependent on how you sequenced your amplicons and the ploidy of your targeted organism.
Assuming you PE sequenced the amplicons then presumably the amplicons were of a length such that the PE insert sizes allowed for internal end overlaps. Internal end overlaps would allow you to discriminate and filter out sequencing artifacts within the body of your amplicons.
The polymorphic variations you report as being in flanking sequences (flanking the SSRs???) could be real reflecting multiple copies of the SSRs with divergent flanks within a single haplotype or between ploidy homologues. Or maybe you have some partially retained amplicon barcodes with more stringent end filtering required.
Much the same comments would apply if you have SE sequenced the amplicons.
Thanks for your answer Stuart
You're quite right, and I have good overlap on most reads, and good depth. Yes, by flanking I meant flanking the repeats.
I have eyeballed some of relevant sets of sequences and I'm thinking the polymorphisms are real, but I'll repeat a subset to confirm.
cheers
Rob
To gain confidence in the SNPs being real, rather than library/sequencing artefacts, then find reads with different starting loci which are overlapping each SNP of interest and check that the SNP still exists in this overlap stack. Using different read start loci for the overlap stack greatly reduces false calls due to PCR differential amplifications of individual fragments. Also look at the background error rates in a window around the putative SNP.
I’m using an extension of the foregoing for my localised haplotype detection method which is giving promising results on allopolyploids..
- Badges
- Science method