#GI2018 - Day Four

Variant Discovery and Genome Assembly

Melissa Wilson Sayres

Sex Differences in Reference Genome Affect Variant Calling and Differential Expression

X and Y homology dotplot

  • a few regions of alignment (PAR)



  • https://www.biorxiv.org/content/early/2018/07/18/346940
  • preprint on correct tech biases on sex chr in NGS data
  • infer sex chromosome complement
  • pointing out regions on autosomes with sex chr - like alignment
    • which is screwing up read balance
    • which presumably could influence variant calling

chrY var calls will have many errors

  • 10,000 variants can be missed
  • mostly in PAR

Prithika Sritharan (Dicks)

Encoding yeast genomic diversity using variation graphs


“one of earliest species to be domesticated”


  • 4,000 strains of 530 species of yeast
  • 1,000 have been sequenced

Want to characterize these

Want to avoid conventional ref genome

  • doesn’t represesent all species well

Variation Graphs

  • using VG Toolkit (Garrison et al. 2017)
  • can align way more to VG than a traditional linear reference
    • looks WAY more, but y axis kind of deceiving (87-99% scale)

Zemin Ning (Durbin)

ScaffHiC Genome Scaffolding by Modelling Distributions of Hi-C Paired-end Reads

Contact matrix

  • C(i,j) num of paired reads between i and j contigs
  • curve fit contact distance with (rougly) quadratic model
  • also use orientation of reads

Quantify contig distance index

difficult to pick ‘favorite’ contig partner

showing stats of icefish genome assmembly which has better n50 etc stats than other tools


comments powered by Disqus