| Abstract Detail
Systematics McKenzie, Patrick [1], Eaton, Deren [2]. Locus length and species network parameters impact gene-tree-based species network inference. Gene-tree-based methods for species network inference are popular tools in plant systematics. These methods infer a set of gene trees from genetic sequences, and they then infer a species tree or network from the set of inferred gene trees. A key assumption of these methods is that the set of inferred gene trees accurately reflects the true distribution of unobservable coalescent histories (genealogies) expected under the species network model. We use phylogenomic simulations with ipcoal to show that the expected length (in base pairs) representing each genealogy along the chromosome depends on species network parameters like branch lengths, effective population sizes, generation times, recombination rate, and the number of tips. When the lengths of loci used for gene tree inference greatly exceed the lengths of the true genealogies, the set of inferred gene trees becomes biased toward the concatenation tree, potentially obscuring or biasing network inference. Careful decision-making could help optimize locus length specification for gene-tree-based methods. Alternatively, SNP-based methods have the appealing property of avoiding locus length specification altogether.
Log in to add this item to your schedule
1 - Columbia University, Ecology, Evolution, and Environmental Biology, 10th Floor Schermerhorn Ext., 1200 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:
Monarda
Systematics
phylogenomics
hybridization.
Presentation Type: Oral Paper
Session: SYSTII, Systematics II: Basal Asterids & Euasterids I
Location: Virtual/Virtual
Date: Tuesday, July 20th, 2021
Time: 2:00 PM(EDT)
Number: SYSTII007
Abstract ID:682
Candidate for Awards:None |
