The study of protein synthesis, and the study of gene expression in general, has accelerated in recent years. Following the advent of next-generation RNA sequencing, powerful library preparation paradigms were developed to capture regulatory activity on a genome-wide scale. In particular, ribosome profiling has emerged as a widely-used measurement of translation. In this method, the state of ribosome association across the transcriptome is obtained by isolation and sequencing of the regions of RNA bound by ribosomes, revealing a snapshot of ribosome positions from which gene-specific densities can be calculated. In combination with RNA sequencing for a measurement of baseline transcription in the same samples, ribosome profiling offers a metric of “translation efficiency”, or TE, corresponding to the average ribosome load per given transcript. Ribosome profiling has advanced the study of translation considerably. However, low throughput in the generation of ribosome profiling and RNA sequencing libraries limits the scale of the experiments that can be performed, while issues in the interpretation of aligned ribosome-protected footprints complicate their analysis, especially in systems of complex regulation. The analysis of such regulatory systems would be greatly aided by a high-throughput sequencing method that can capture translational regulation, but current methods of measuring genome-wide translation are inherently limited in scale.
This thesis addresses the key issues presented above in separate chapters. Chapter 2 discusses the analysis of elongation and initiation from ribosome profiling and RNA sequencing data in a mouse model of Fragile X Syndrome. In this chapter, several methods of measuring and modeling variability in the distribution of ribosomes along a coding sequence are used alongside analyses of differential RPF and RNA abundances and their ratio, RFApm, which we distinguish from TE to emphasize its dependence on factors other than initiation rate. The chapter summarizes current information regarding the observed effects of FMRP, and proposes a model congruent with these observations and more-recently published studies. Chapters 3 and 4 present approaches to modeling or inferring translational regulatory networks, either by a novel library preparation paradigm or computational inference from publicly-available data. Chapter 3 presents riboPLATE-seq, a high-throughput RNA-seq library construction method based on the existing PLATE-seq method. The method recapitulates significant findings from ribosome profiling and RNA sequencing at a fraction of the per-sample cost, with further advantages in scalability, and could be implemented in a large-scale screen of translational regulators to create a network of their specific targets. Chapter 4 presents an approach to inferring translational regulation from integrative analysis of public ribosome profiling and RNA sequencing data, tailoring the powerful inference engine ARACNe to measure translational interactions. This yields a comprehensive network of translational regulation, assigning target genes to the set of RNA-binding proteins.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/9f4r-e104 |
| Date | January 2022 |
| Creators | Metz, Jordan Benjamin |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0025 seconds