Seminars

May 17 2018

5:00 pm 1100 TLSB

EcoEvoPub Series

Graduate Student Presentations

Summary

Tina Del Carpio

"mSTARR-Seq: Genome-Wide Assay for the Effect of DNA Methylation on Enhancer Activity"

DNA methylation is a key epigenetic mark that targets CG di-nucleotides (CpG sites) and can modulate the expression of RNA without altering the underlying DNA sequence. However, the magnitude and directionality of DNA methylations impact on gene expression, if any, is site-specific. While there is growing literature supporting the connection between DNA methylation, aging, and disease state, tools to study its functional effects on a genome-wide scale are limited. We describe a new method, mSTARR-seq, that allows for high-throughput testing of DNA methylations effect on gene expression in a cellular context. mSTARR-seq has been adapted from STARR-seq, a method used to assay enhancer activity on a genome-wide scale. However, the mSTARR-seq plasmid construct is CpG free allowing for scaling up of gene expression reporter assays to test tens to hundreds of thousands of regions of the genome at one time, as well as the effects of unmethylated and methylated versions of the same locus. We report the results of applying mSTARR-seq to assay millions of CpG sites in the K562 human myeloid lineage cell line. We identify thousands of sites where methylation state impacts expression levels. Furthermore, we find that identity of transcription factor binding sequences and chromatin state of a given region can partially predict when methylation state impacts expression. Our results provide a broad survey of sites functionally affected by DNA methylation in the human genome. mSTARR-seq is a generalizable approach that can be applied in diverse cell types and species to study the effects of DNA methylation.


Tyler McCraney

"Phylogenomic Systematics of Gobiiformes: Small Fishes Show Some Backbone"

Gobiiformes comprise a global radiation of small, mostly marine fishes characterized by extraordinary species richness and ecological diversity. Molecular phylogenetics has been crucial to advancing gobiiform systematics, but classification remains unsettled due to disagreements on clade relationships and poor bootstrap support\; especially along the backbone of the tree subtending lineages of goby families Gobionellidae and Gobiidae. Here I use a phylogenomic approach to infer gobiiform phylogeny from 855 ultraconserved element loci sequenced on a broadly sampled set of 50 taxa. Consistent with early mitochondrial work and recent phylogenomic studies, my results support successive sister group relationships of families Kurtidae, Apogonidae, Odontobutidae, Eleotridae, Butidae, Gobionellidae and Gobiidae. I also recover monophyly of all goby lineages sampled with complete bootstrap support. These results support the classification of Gobiiformes inclusive of Kurtiformes (Kurtidae and Apogonidae), provide improved resolution along the gobioid backbone, and advance systematics of goby lineages.


Thursday, May 17, 2018
5:00pm
1100 TLSB

 

 

 

 

 

 

 

 

 

 



 

 

 

 

 

 

 

 

 

 

 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



 



this is idtest: