search

zhengxwen / SeqArray

Data management of large-scale whole-genome sequence variant calls (Development version only)
http://www.bioconductor.org/packages/SeqArray
44 stars 12 forks source link

seqGDS2SNP() altering genotype values for haploid data #42

Open AAvalos82 opened 5 years ago

AAvalos82 commented 5 years ago

Came across an issue where using seqGDS2SNP() on a "SeqVarGDSClass" object housing haploid variant calls that had been filtered. Effectively the process alters the genotypes, with the resulting "SNPGDSFileClass" object's genotype matrix not matching at all to the source data (even missing values are replaced). Extracting the genotype array from the source "SeqVarGDSClass" object works fine, and so does extraction of the dosage matrix, both of these match each other (with the dosage value accounting for REF allele count as it should). However it seems that when seqGDS2SNP() parses through the "SeqVarGDSClass" genotype array it produces something else altogether.

I also tried to produce the "SNPGDSFileClass" object with the dosage = T option, and this resulted in:

Error in seqApply(gdsfile, dosage, as.is = gGeno, .progress = verbose, : The GDS node "annotation/format/DS/data" does not exist.

This may be an unrelated issue, but I thought it would be worth noting.

complexgenome commented 3 years ago

Were you able to resolve this?