Presentation Detail

Molecular Evolution

Ding, Yuantong [1], Wu, Steven [2], Sukumaran, Jeet [3], McConnell, Christie  [4], Rodrigo, Allen [4].

Evolutionary analyses using reconstructed haplotypes from next-generation sequencing of mixed samples: an evaluation of current programs.

While next-generation sequencing (NGS) technology provides large volumes of sequence data at lower cost compared to other conventional methods, it also brings new challenges in analyses due to the length of short reads. In phylogenetic or population genetic analysis, one way to use these short reads is to first assemblethe fragments into longer sequences. Then the standard tree-reconstruction analyses are applied to these reconstructed sequences. To test how well this method can infer the actual phylogeny and evolutionary information, this study uses simulations to evaluate the performance of different haplotype reconstruction software. We simulated genealogies of full-length sequences using a coalescent model and short read data with 454-error model. Comparing the phylogenetic tree and population parameters estimated based on the reconstructed haplotype to the true value revealed that short reads reassembling methods may give misleading results. Software should be carefully chosen based on the data used and in some cases, flittering the reconstructed haplotypes according to the phylogenetic diversity may help increase the accuracy.

1 - Duke University, Biology, Durham, NC, USA
2 - Duke University, Biology Department, 130 Science Drive, Biological Science Building, Duke box 90338, Durham, NC, 27708, United States
3 - Duke University, Biology, Rm 137 Bio Sci, Box 90338, 130 Science Drive, Duke University, Durham, NC, 27708-0338, USA
4 -

Keywords:
haplotypes reconstruction
next-generation sequencing
quasispecies.

Presentation Type: Regular Oral Presentation
Session: 113
Location: Alpine A and B/Snowbird Center
Date: Monday, June 24th, 2013
Time: 11:30 AM
Number: 113005
Abstract ID:905
Candidate for Awards:Ernst Mayr Award