Create your own conference schedule! Click here for full instructions

Abstract Detail


Folk, Ryan [1], Allen, Julie [1], Soltis, Pamela S. [2], Soltis, Douglas E. [2], Guralnick, Robert [1].

Data Assembly and Post-Processing in aTRAM for Museum Phylogenomics.

NGS methods have quickly become a central resource for phylogenomics, enabling robust phylogenetic inference in a wide variety of taxa and greatly facilitating the use of historical DNAs. Methods for data collection and multilocus phylogenetic analysis are now well-trodden territory, but in the middle ground there is a conspicuous literature gap of identifying appropriate methodologies for data assembly. Much has been proposed in the way of alternative methods, but little exists in the way of comparative methodological studies. Several pipelines have appeared, among which aTRAM is particularly promising.
aTRAM is a hybrid reference-based and reduced de novo method (that is, a reference-guided method). Reads matching a reference by BLAST are pulled and assembled as small de novo subproblems; this process is iterative in the sense that assembly results are the new reference for each successive round of BLASTing and assembly. aTRAM is highly parallelizable within each assembly problem and among loci and taxa, and is being actively optimized for performance and ported to python; the de novo portion integrates multiple methods with highly flexible parameters. Here, we report aTRAM results for (1) a deep divergence, comprising exon-capture data collected in the Saxifragales. This late Cretaceous radiation presents an assembly challenge due to deep nucleotide divergence and our reliance on primarily museum DNAs. We also show (2) results for a shallow divergence in Heuchera, a ~4 my divergence where low nucleotide variation and the acute need for gene-tree based analyses present the primary challenges. We compare these toreference-based short read aligning approaches and present new work in post-processing exonic data in a protein- and paralog-aware manner, complementary to alternative consensus-based approaches. We advocate a pluralistic approach to NGS assembly; much like de novo assembly, researchers should begin by testing several methods, and perhaps report comparative results among assemblies, which could shed light on methodological or biological issues that would be hidden in isolation.

Log in to add this item to your schedule

1 - University of Florida, Florida Museum of Natural History, Dickinson Hall, 1659 Museum Road, Gainesville, FL, 32611, USA
2 - University Of Florida, Florida Museum Of Natural History, PO BOX 117800, Gainesville, FL, 32611-7800, USA, 352/273-1964


Presentation Type: Oral Paper
Session: 29, Phylogenomics I
Location: Fort Worth Ballroom 4/Omni Hotel
Date: Tuesday, June 27th, 2017
Time: 1:30 PM
Number: 29001
Abstract ID:28
Candidate for Awards:None

Copyright © 2000-2017, Botanical Society of America. All rights reserved