R01: A kinetic framework to map the genetic determinants of alternative RNA isoform expression Project Summary At the core of the central dogma is the transcription of RNA, which enables instructions from DNA to be translated into protein messages. The biogenesis of mRNA is a complex and highly regulated process, re- quiring coordination between transcriptional and RNA processing machinery that each comprise hundreds of regulatory RNAs and proteins. These interactions often result in many possible alternative isoforms ex- pressed from a single gene. While we can now extensively catalog the abundance and variability of mRNA isoforms across cellular states, we are still limited in our abilities to predict coordination between the mech- anistic steps that give rise to cellular diversity. Most catalogs of mRNA levels only profile steady-state mRNA and are blind to the spatiotemporal dynamics regulating isoform choice and expression. However, the trajectory and fate of an mRNA molecule likely depends on the efficiency of kinetic interactions at each step of mRNA biogenesis. Thus, we must directly track mRNA fluxes on timescales commensurate with the regulatory decisions being made to estimate the rates of each step of mRNA biogenesis. While high-throughput approaches to quantify steady-state RNA levels have matured in the past decade, it is still challenging to quantify nascent RNA and infer the rates of core mRNA biogenesis steps including transcription initiation, elongation, splicing, and 3’ end cleavage. Both technical and mechanistic features confound direct kinetic measurements from nascent RNA sequencing data. Existing approaches to quan- tify rates of these steps have found extensive variability in these rates across genes and cell types, but have been limited in their ability to estimate rates of alternative processing decisions, identify kinetic coordina- tion between splicing or 3’ end cleavage events, and map genetic elements that underlie kinetic variability. We propose to address these challenges by developing a joint experimental and statistical framework to characterize the molecular and genetic factors affecting RNA processing rates. First, we will build a statistical model to estimate RNA processing rates at individual sites, and generate time-resolved short-read nascent RNA-sequencing data to train this model. Second, we will characterize ki- netic competition between individual events that underlies alternative isoform expression using a constrained state space model and generate long-read RNA sequencing data to identify expressed isoform trajectories. Third, we will estimate RNA processing rates in a population of genotyped human cells to identify quan- titative trait loci (QTL) associated with variability in RNA processing rates. For all three aims, we identify rigorous validation metrics and calibrate the uncertainty in our rate predictions. Together, our work will lead to i) a better understanding of how molecular coordination and genetic variants regulate the rates and deci- sion points in RNA processing, ii) a generalizable genomics framework for quantifying these rates, and iii) a QTL resource to interpret human disease-associated genotypes acting through rate-distorting mechanisms.

The biogenesis of mRNA is a complex and highly regulated process, requiring coordination between transcrip- tional and RNA processing machinery that each comprise of hundreds of regulatory RNAs and proteins, resulting in many possible alternative isoforms expressed from a single gene. While we can now extensively catalog the abundance and variability of mRNA isoforms across cellular states, we are still limited in our abilities to predict coordination between the molecular steps that give rise to cellular diversity and the rates at which each of these mechanisms occurs. Here, we propose to generate checkpointed time-series RNA-sequencing data and robust statistical models to quantify RNA processing rates, fully characterize the kinetic competition underlying alterna- tive isoform expression, and identify genetic variants associated with variabilty in RNA processing rates across human cell lines.