Genomic Analysis of Non-model and Emerging Model Organisms
Reed A. Cartwright, PhD
Arizona State University
Assistant Professor of Genomics, Evolution, and Bioinformatics
School of Life Sciences
Human and Comparative Genomics Laboratory
The Biodesign Institute
Advancements in sequencing technology and bioinformatic infrastructure have generated a deluge of genomic data across the tree of life. However, artifacts are a common problem in genomic datasets, increasing curation efforts and causing a large amount of data to be discarded. Importantly, uncorrected genomic errors lead to bad alignments, erroneous estimates of important evolutionary parameters, and other problems with comparative and functional genomic studies. In response, we are developing methods that use advanced statistical models to capture the biological and technical processes that shape genomic data. By robustly modeling the biological and technical process that shape genomic data, we aim to properly handle non-model and emerging model organisms. Our methods are being implemented in open-source and user-friendly applications using the best practices for software engineering.
In the first part of this presentation, I will describe our progress developing DeNovoGear and MUTK, two software packages designed for the estimation of mutation rates from related individuals and samples. I will describe our application of DeNovoGear/MUTK to the estimation of somatic mutation rates in a long-lived Eucalyptus tree, including answering the long standing question of whether it is possible to reconstruct the physical topology of a living tree using only somatic mutations.
In the second part of this presentation, I will describe our progress on developing COATi, a multiple-sequence alignment tool to improve the analysis of coding sequences from non-model and emerging model organisms. COATi will enable users to generate accurate alignments for comparative and functional genomic studies. These methods will combine cutting-edge models of protein coding-sequence evolution --- including mutation-selection models --- with robust models of sequence-assembly errors. Although this project just started, early implementations are producing better pairwise alignments than existing options.
Dr. Reed Cartwright is an Assistant Professor in the School of Life Sciences and the Biodesign Institute at Arizona State University. He is currently leading the Software Carpentry initiative at ASU for modern pedagogy in scientific computing and data analysis. In his research, he uses computational and statistical approaches to study evolutionary processes using genomic data. He works on a variety of species including model, non-model, and emerging model organisms. Most of his research involves developing and applying novel software to study mutation processes across the tree of life.