Haplotyping for different sample types (individual or Pool-Seq)

Individual and Pool-Seq samples are analyzed differently in three important ways;

• In individuals, low frequency haplotypes (<5%) are generally noise and need to be removed from the data. In Pool-Seq, it may be difficult to discriminate between noise and low frequency true alleles.

• In Pool-Seq, haplotype frequencies represent the genetic diversity of the population. For individuals, haplotype frequencies should be converted into discrete genotype calls, which can either be expressed as dominant (0/1) or dosage (0/1/2, diploid; 0/1/2/3/4, tetraploid).

• In individuals, the total number of haplotypes (sum of dosage calls per locus per sample) can be compared to the number of alleles present in the nuclear genome (2 for diploids, 4 for tetraploids). This provides to opportunity to remove loci with consistent bad calls across a large fraction of the sample set, or to identify samples with substantial numbers of loci with unexpected dosage calls (thus identifying samples with mis-labeled ploidy levels, or contaminated DNA samples).

Therefore, these two sample types are discussed separately. Tables with haplotype frequencies are always calculated, while specific options are available to perform discrete genotype calling and subsequent haplotype call filtering based on expected dosage calls in diploid or tetraploid individuals.

Contents:

• Individuals
  • Step 4: Filtering haplotypes and calculating haplotype frequencies
  • Step 5: Transforming haplotype frequencies to discrete calls in individuals
  • Step 6: Filtering for high quality dosage calls in individuals
  • Output
  • Summary of Commands
• Pools
  • Step 4: Filtering haplotypes and calculating haplotype frequencies
  • Output
  • Summary of Commands