I’m trying to make a genetic map with SLAF-seq to a F1 mapping population of a perennial woody plant. SLAF-seq is a genotyping by sequencing technique similar to GBS or ddRAD. But I now find almost 70% of the F1 offsprings have ~2%(1000-4000) loci that are not the same as corresponding loci of any of the parents. The average sequence depth is ~4X.
For example, one locus is as follows:
ACAACCCAAGAACAAATAACGTTTATTTACACATGTTTCTTCAATACATCGGAGCCGCTCTTGTATTAGTCAATAAAAATXXXXXXXXXXATAGTTAAAAACTATCCGCCACGATCAAAAAACAGAGCTTTTGTTTCACCAGGTAGTCTCGCTGCACGGGTAGTGCTAGT ab（progeny）,
ACAACCCAAGAACAAATAACGTTTATTTACACATGTTTCTTCAATACATCGGAGCCGCTCTTGTATTAGTCAATAAAAATXXXXXXXXXXATAGTTAAAACCTATCCGCCACGATCAAAAAACAGAGCTTTTGTTTCACCAGGTAGTCTCGCTGCACGGGTAGTGCTAGT M,P（parents）
Then I began to doubt that the parents were polluted by pollen of other plants, but then this should not go wrong in so large a scale for the strict control.
Is anyone here familiar with this technique? Is it mutation or sequencing errors or something else? Is it normal for such a proportion? Does anyone meet similar situation in dealing with GBS or ddRAD data?
Your answers will be appreciated. Best regards.
SLAF-seq ref：http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0058700
It could be that you are not identifying the full parental genotype at the 2% of loci. A heterozygous parental locus is called homozygous, but the uncalled allele then is seen in the F1s. If the sequencing depth in the parent libraries is 4X, then I am surprised you don't have more of these alleles appearing, since sampling two chromosomes 4 times gives you a 0.5^4 chance of sampling just one of the chromosomes in the 4 reads.
The usual causes of allele dropout in GBS/ddRAD data (different size selections, polymorphism in a diverse population affects a cut site) wouldn't apply here.
Thanks very much for your answer, Dr Johnson. I am sorry for my unclear expression. In fact the sequencing depth in the parent libraries is deeper, about 20X for each parent and the average sequence depth of the progenies is ~4X. Is it still possible that we have not identified the full parental genotype at the 2% of loci? What level of allele dropout in GBS/ddRAD data is reasonable for this sequencing depth? In our experiment a 4 bases restriction enzyme was chosen to cut the genome and ~150,000 tags were produced. The genome is ~2G and we sampled fragments only between 550-600bp.
- Badges
- Science method