Skip to content

Instantly share code, notes, and snippets.

@caseywdunn
Last active Oct 30, 2016
Embed
What would you like to do?
Notes from my October 7, 2016 talk at the Boston Evolutionary Genomics Supergroup Meeting

Boston Evolutionary Genomics Supergroup Meeting

October 7, 2016

Casey Dunn, http://dunnlab.org , @caseywdunn

These are some notes and annotated references from my talk at the Boston Evolutionary Genomics Supergroup Meeting.

Papers and analyses from my research group that I mentioned today

  • Dunn, CW, C Munro (2016) Comparative genomics and the diversity of life. Zoologica Scripta 45:5-13. doi:10.1111/zsc.12211.

This paper touches on the false dichotomy between "descriptive" and "hypothesis driven" research projects. It also describes a few other challenges in evolutionary genomics, including the diminishing usefulness of orthology and paralogy as well as biases that may be introduced in analyses focused on single copy genes.

A critique of Levin et al.'s conclusion that the origin of animal phyla was associated with distinctive changes in develomental gene expression.

A reanalysis of Levin et al. that illustrates how evolutionary changes in expression that are specific to a single lineage can be misinterpreted as a general pattern when making pairwise comparisons between species rather than using phylogenetic reconstruction methods.

  • Dunn, CW, X Luo, Z Wu (2013) Phylogenetic analysis of gene expression. Integrative and Comparative Biology 53:847-856. doi:10.1093/icb/ict068

Challenges and approaches for phylogenetic analysis of gene expression. It isn't statistically valid to map RNA-seq counts onto a phylogeny, but certain ratios of counts can be reconstructed on trees.

  • Guang, A, F Zapata, M Howison, CE Lawrence, CW Dunn (2016) An Integrated Perspective on Phylogenetic Workflows. Trends in Ecology and Evolution 31:116-126. doi:10.1016/j.tree.2015.12.007.

A description of the implicit statistical assumptions, including low relative entropy, that are made by most current phylogenetic (and functional genomic) analysis workflows. There are alternative engineering approaches that could be made that would relax these assumptions.

  • Dunn, CW, SP Leys, SHD Haddock (2015) The hidden biology of sponges and ctenophores. Trends in Ecology and Evolution 30:282-291. doi:10.1016/j.tree.2015.03.003.

The impact of ascertainment biases on our understanding of diversity.

Other resources

  • Breiman, L (2001) Statistical Modeling: The Two Cultures (with comments and a rejoinder by the author). Statistical Science 16:199-231. doi:10.1214/ss/1009213726.

A very interesting take that is quite different than the one I present on how useful model based approaches are. tl;dr: For many predictive tasks, models are neither necessary nor desirable. The rejoinders argue that models are still critical when asking questions about unknown mechanisms.

  • Levin M, et al. (2016) The mid-developmental transition and the evolution of animal body plans. Nature 531:637–641. doi:10.1038/nature16994.

The authors collected an impressive time series of gene expression data through embryonic development for 10 distantly related animals. I have several concerns with the interpretation and implementation of the analyses.

Footer

This document can be accessed via the delightful github url shortener at https://git.io/genomics2016 .

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment