Analyzing single cell pollen sequencing with MIGseq libraries
Our goal is to genotype single pollen grains - better than single cell because pollen is hapolid and has 3 cells each - to detect the location of recombination. This project is a work in progress.
MIGseq uses degnerate primers in a multiplexed PCR reaction to amplify and sequence ISSR (inter-simple sequence repeat) markers by the hundreds or thousands. I modified the MIGseq library prep of Suyama and Matsuki, 2015 to include Adapterama I barcodes (Glenn et al, 2019a) and Adapterama II style primers (Glenn et al, 2019b).
We did a power analysis to determine how many pollen grains we would have to sequence. Imputing missing data for pollen is easy if we assume 1 or 2 crossover events per chromosome because pollen is hapolid.
- Assume less recomb near telomeres/centromeres: cosine distribution
- simulate parents as just 0 or 1 (2 alleles)
- sample games as combination of 0, 1
- recode gametes to represent parents
- Draw gamete haplotypes
- infer recomb rate between each SNP position