Tweet
Hi
I am currently analysing a newly Illumina sequenced invertebrate genome which is highly related to a publicly available reference genome (only trace files available). Both species are thought to reproduce primarily vegetative. The new genome was thought to represent a different species based on mitochondrial 16S. The whole genome alignments show high identity (99%) and we can even map the Illumina reads of the new genome to the reference (resulting in a higher SNP rate and fewer reads mapping). If I look by eye on particular loci I can see that both genomes always seem to share one allele and the other one is different, suggesting that both rather represent a species complex and one is a parent the other a hybrid. But what would be the appropriate analyses to prove that?
My first approach was to produce a phased vcf-file from variant calling (samtools pipeline: the trace files to the reference, the Illumina PE to the new genome) using HapCut and the tool HapCutToVcf from Fgbio. From that I wanted to generate blocks of phased alleles as fasta sequences and blast them against the phased allele blocks of the other genome. If a great portion of the phased blocks align with 100% identity to the other phased blocks then it seems likely that they share 50% of alleles. Unfortunately I only found one tool (vcfx) which claims to generate fasta from phased vcf and this did not work with my phased vcf.
Even if that worked, a problem with that approach would be the quantification, because even if there was no further sex (further crossovers) after split of the two lineages, the size of the phased blocks between the two samples will differ, simply due to the different experimental designs of the sequencing.
So my questions are: 1. Are there other, or better, ways to prove that (roughly) 50% of the alleles are identical (one is likely a hybrid)? 2. Are there other tools which can generate fasta from phased vcf?
Thanks!
I am currently analysing a newly Illumina sequenced invertebrate genome which is highly related to a publicly available reference genome (only trace files available). Both species are thought to reproduce primarily vegetative. The new genome was thought to represent a different species based on mitochondrial 16S. The whole genome alignments show high identity (99%) and we can even map the Illumina reads of the new genome to the reference (resulting in a higher SNP rate and fewer reads mapping). If I look by eye on particular loci I can see that both genomes always seem to share one allele and the other one is different, suggesting that both rather represent a species complex and one is a parent the other a hybrid. But what would be the appropriate analyses to prove that?
My first approach was to produce a phased vcf-file from variant calling (samtools pipeline: the trace files to the reference, the Illumina PE to the new genome) using HapCut and the tool HapCutToVcf from Fgbio. From that I wanted to generate blocks of phased alleles as fasta sequences and blast them against the phased allele blocks of the other genome. If a great portion of the phased blocks align with 100% identity to the other phased blocks then it seems likely that they share 50% of alleles. Unfortunately I only found one tool (vcfx) which claims to generate fasta from phased vcf and this did not work with my phased vcf.
Even if that worked, a problem with that approach would be the quantification, because even if there was no further sex (further crossovers) after split of the two lineages, the size of the phased blocks between the two samples will differ, simply due to the different experimental designs of the sequencing.
So my questions are: 1. Are there other, or better, ways to prove that (roughly) 50% of the alleles are identical (one is likely a hybrid)? 2. Are there other tools which can generate fasta from phased vcf?
Thanks!