When do autopolyploids need poly‐sequencing data?

Abstract

The sequencing depth required to genotype autopolyploid populations is a very controversial topic. Different studies have adopted variable depth values without a clear guide on the optimal sequencing depth value. Many studies suggest high depth thresholds for different ploidies that may not be practical and substantially increase the overall genotyping cost for different projects. However, such conservative thresholds may not be required to achieve the most common research goals. In fact, some recent reports in the field of quantitative genetics found that much lower sequencing depth thresholds could achieve the same accuracy as high depth thresholds. In this manuscript, I discuss when researchers need to use stringent sequencing depth thresholds and when they can use more relaxed ones. I support my argument by calculating the probabilities of sampling different homologues at a given sequencing depth. I also discuss the uses and the uncertainty in calculating a continuous allelic dosage as the proportion of sequencing reads that hold the alternative allele, which is becoming a common method now in quantitative genetics to replace discrete dosage estimation.

Construction of relatedness matrices in autopolyploid populations using low-depth high-throughput sequencing data

“…An important consideration for researchers planning to conduct HTS studies on polyploid species is the average read depth to aim for when sequencing. Sequencing depth in autopolyploid populations is a controversial topic and highly stringent deep sequencing often recommended in the literature can be cost prohibitive and, depending on the application, unnecessary if the uncertainty in the dosage information is accounted for in the analysis (Jighly 2022 ). For a fixed amount of sequencing resources available, there is a trade-off between the number of SNPs called and the average read depth at a SNP.…”

Variant calling in polyploids for population and quantitative genetics

“…This disparity in sequencing cost at low coverage is increased by many existing polyploid genotyping algorithms requiring high coverage to overcome allelic dosage uncertainty, which is the ambiguity in the number of alternate allele copies in polyploid genotypes . The minimum coverage requirement to obtain high-confidence genotypes may range from 10 to over 50X depending on the ploidy level and genotyping software, whereas diploids need only 8X coverage Jighly, 2022) . After sequencing has been accomplished, access to HPC is needed for data storage and analysis because the size of sequence alignment files (BAMs) and variant call files (VCFs) produced in the variant calling pipeline scale with genome size and sample size (Muir et al, 2016;Weiß et al, 2018) .…”

“…After sequencing has been accomplished, access to HPC is needed for data storage and analysis because the size of sequence alignment files (BAMs) and variant call files (VCFs) produced in the variant calling pipeline scale with genome size and sample size (Muir et al, 2016;Weiß et al, 2018) . Failing to sequence to sufficient coverage or limiting sample size to meet budget constraints may result in insufficient sampling of alleles and rare variants, the misestimation of allele frequencies, and low power in analyses like admixture analysis and genome wide association (Jighly, 2022) .…”

Genotyping by Sequencing Advancements in Barley

“…Among cereals, wheat (16 GB) and barley (5.5 GB) have large complex genomes with over 80% of repetitive elements ( Wicker et al, 2009 ). Hence, sequencing has not always been a straightforward approach for these crops, and thus a higher sequencing depth is required to achieve SNP calling ( Dou et al, 2012 ; Jighly, 2022 ). To overcome these challenges in SNP identification, the genome size reduction was introduced as a low-depth sequencing strategy ( Huang et al, 2009 ; Scheben et al, 2017 ).…”