Hi ,

I used seqVCFtoGDS to convert VCF into GDS (My VCF files available per chromosomes are further divided into chunks e.g. chr1.1.vcf.gz, chr1.2.vcf.gz ... converted all chunks into single chr1.gds for chromosome 1 likewise have 22 gds files).

Is this the correct way to convert vcf to gds per chromosome wise?

For calculating PC-AiR and PC-Relate SeqVarData need to have all chromosomes information in single GDS file ,or is it possible to use chromosome wise GDS files for the analysis without combining?

Regards, JT

Your conversion strategy is indeed what I would recommend, and having separate files for each chromosome is helpful to speed up association testing. For PC-AiR and PC-Relate you do need to have all variants in a single file, but I recommend LD pruning (with the SNPRelate function snpgdsLDpruning) to select variants. Instead of merging all variants into a single file, you could run LD pruning on each of your per-chromosome GDS files, then merge only selected variants into a single file with seqMerge.