UCLA

A majority of the work presented in this dissertation focuses on identifying differences in transcriptome profiles across different phenotypes. The first project I present incorporates controls from different developmental time points, namely, prenatal and postnatal, to identify gene expression and splicing differences in SMA cases. Findings from this study report a large number of genes with prenatal expression patterns in iliopsoas from postnatal SMA samples. Similarly, differential splicing analyses uncovered prenatal splicing patterns in SMA cases in two muscle relevant genes: TNNT3 and MYBPC1.

The next project characterizes the transcriptome profile of seven different tissues in the vervet monkey using RNA-seq data. Transcriptome profiles from two of the three brain tissues explored showed expression patterns correlated with developmental time point. Additionally, this project presents an eQTL study which resulted in identifying eQTL SNPs within a region associated with hippocampal volume.

Building on the observation of developmental expression patterns in Brodmann’s area 46 and caudate in the previous project, the next project I present focuses on the identification of age-related genes in vervet hippocampus. With the addition of younger samples, I also perform an eQTL analysis and report two additional genes, CHMP1B and RAB31, with associated SNPs within the hippocampal volume associated region.

Finally, the final project described focuses on improving the characterization of vervet chromatin modifications using human epigenomic datasets. Through the use of machine learning algorithms and prediction variables previously shown to correlate with conversion depth of histone marks across species, I show improved accuracy can be obtained while still maintaining biologically relevant peak signals.