Create your own conference schedule! Click here for full instructions

The Virtual Conference is located at https://botany2021.pathable.co/.

Abstract Detail



Phylogenomics

Cohen, Rachel [1], Eaton, Deren [2].

Hogtie: A Reference-Free Tool for Identifying Introgression in Plant Genomes.

In the genomics era, likelihood-based models of discrete character evolution are used to compare genomic datasets under different models of evolution. Such models can be used to identify genes that are more similar to divergent lineages than to closely related lineages; for example, genes gained through introgression, horizontal gene transfer, or convergence. These methods tend to rely on a priori homology identification, which requires mapping reads to well-resolved reference genomes. Identifying homology in this way can bias comparisons with regard to the reference genome itself. Reference-based biases are particularly problematic in plants, which are highly variable in genome size and structure. K-mer based methods have become popular for reference-free comparative genomics, but are designed to compare only two genomes at once and do not account for phylogenetic relationships. To address these shortcomings, we introduce the python program hogtie. Hogtie treats k-mer presence/absence as a binary character and models evolution backwards-in-time via a discrete Markov model, flagging k-mers whose presence/absence patterns deviate from expectations. Expectations are set by parameterizing a provided species tree through evolutionary simulations, a flexible approach that can take into account diverse evolutionary scenarios (i.e. high ILS or slow mutation rates). The hogtie methodology is particularly effective because it combines the advantages of traditional models of discrete character evolution and reference-free analyses, providing a tree-based analysis that does not require genome assemblies for all species included in analyses. In addition, the framework of discretizing sequence data can be applied beyond k-mers to SNPs or transcripts, making hogtie's method broadly applicable.  


Log in to add this item to your schedule

Related Links:
Github Repository for Hogtie


1 - Columbia University, 1190 Amsterdam Ave, New York, NY, 10027, United States
2 - Columbia University, Ecology, Evolution, And Environmental Biology, 1200 Amsterdam Ave. , Schermerhorn Ext. Office 1007, New York, NY, 10027, United States

Keywords:
modeling
phylogenetics
K-mer
HGT
Introgression
transcriptomics.

Presentation Type: Poster
Session: P3, Phylogenomics Posters
Location: Virtual/Virtual
Date: Wednesday, July 21st, 2021
Time: 5:00 PM(EDT)
Number: P3PL020
Abstract ID:658
Candidate for Awards:None


Copyright © 2000-2021, Botanical Society of America. All rights reserved

aws4