Computer Science Distinguished Lecture
| Date/Time: | Thursday, 28 Mar 2013 at 3:40 pm |
|---|---|
| Location: | 207 Marston Hall |
| Cost: | Free |
| Phone: | 515-294-6516 |
| Channel: | College of Liberal Arts and Sciences |
| Categories: | Lectures |
| Actions: | Download iCal/vCal | Email Reminder |
Abstract
In many applications of NGS, sequenced samples contain multiple distinct but sometimes very similar sequences. This talk addresses two such applications: reconstructing viral quasispecies from shotgun and amplicon reads and transcriptome reconstruction from RNA-Seq single and paired reads.
Error-prone replication of RNA viruses with high mutation rate creates a diverse population of closely related variants known as quasispecies. By understanding the quasispecies, more effective drugs and vaccines can be manufactured as well as cost-saving metrics for infected patients implemented. Reconstructing the quasispecies spectrum is difficult since conserved regions in the genome can extend beyond the read length. We will compare
recent approaches of quasispecies reconstruction from shotgun and predefined amplicon reads. Transcriptome reconstruction and quantification from RNA-Seq reads can be greatly improved by using existing partial annotation as well as fully utilizing paired reads. We describe a framework which removes reads that can be explained by existing annotation and focuses on remaining reads for discovering novel transcripts and estimating its frequency. We next describe optimization approaches minimizing number of candidate transcripts requiring mapped paired reads to closely follow known fragment length distribution and uniformly cover transcripts.