This project is to assess single-site versus multi-allelic variant detection models for site-directed next-gen read sequences. Another word, can we accurately infer genotype variants from next-generation read sequences?
Currently, this is being tested against SAMtools, ANGSD, GATK and FreeBayes. I need to check out OCTOPUS.
Python main/PopMs.py
: used to simulate single-end, medium-size (>300 bp) site-specific illumina pooled sample reads (Fastq file), assuming all sampled diploid individuals drawn from the same population
Disclaimer! My read simulator:
- is only suitable for representing reads from site-directed read library where all reads of a single loci begin on the same site
- does not consider somatic mutation or the event of DNA structural rearrangement
- assumes that individuals are diploid organism
- assumes uniform variant rate across all loci