These are chat archives for dereneaton/ipyrad

Aug 2017
Aug 01 2017 07:59
I knew why my loci and snp are so low,because some samples have relatively high error rate and heterozygou rate,so in the step 5,these sampls keep much less consensus reads than others,is this right?But how do you calculate the heterozygou rate?
there is no publication related to the method of calculating heterozygou rate
Deren Eaton
Aug 01 2017 14:25
Hi @Wind-ant, the original pyRAD publication has a citation for the maximum likelihood equation to jointly estimate heterozygosity and error-rate that is used in step 4. Those estimated parameters are used in step 5 to make base calls, and the heterozygosity rate reported in the step 5 stats output is simply the proportion of sites that are heterozygous (nhetero / nsites).
There are many reasons why the number of loci and snps might be low. It could be because the parameter settings in ipyrad are not correct for your data set. But most of the time low amounts of data are because of the library preparation method that was used, which either selected too few fragments, or too short of fragments, or too many fragments, and this often coincides with too low of sequencing depth.