A framework for systematic promoter motif discovery and expression profiling from high dimensional brain transcriptome data

A framework for systematic promoter motif discovery and expression profiling from high dimensional brain transcriptome data Jeremy A. Lieberman Understanding the regulatory logic of genes across discrete brain substructures can elucidate the basis for neural network connectivity and the cause of disease. Promoter motifs, in particular, that govern high or low expression gene networks present an important fulcrum for phenotypic behavior. Using the Allen Institute Brain Atlas we took various clustering approaches to find closely regulated genes, and generated substructure specific expression profiles to run through FIRE, a motif discovery algorithm and iPAGE, a functional ontology algorithm. Notably, we found a single large cluster of genes that had tightly coordinated behavior across hundreds of brain substructures, as well as a unique upstream promoter signature, yet highly diverse ontological characteristics. We also present a BRain EXpression Profile ASSembly script (BEXPASS) whose output is customized for FIRE and iPAGE input. Lastly we look at language processing and speech control areas of the brain and put forward recommendations for promoters that can serve as part of DNA constructs for optogenetic research an emerging neuroscientific research method that uses bacterial light-gated ion channel protein, channelrhodopsin (ChR1 or ChR2), as an activity control tool to activate neural pathway signaling. ACKNOWLEDGEMENTS I would like to thank the members of Professor Saeed Tavazoie’s lab for supporting my research. In particular, Saeed Tavazoie and Panos Oikonomou for their mentorship and guidance. Also to Peter Freddolino for his expertise of the R language.