Large scale ancestral reconstruction of protein sequence, structure and molecular function

点击翻译按钮获取中文摘要

Using statistical models to infer the sequences of ancestral molecules allows researchers to directly study ancient molecules in the laboratory. These studies have greatly improved our understanding of how molecules work and are being used to generate new therapeutics and inform biochemical engineering efforts. However, errors made when inferring ancestral sequences could undermine downstream laboratory studies. The goal of this project is to characterize how and why ancestral sequence reconstruction might fail and to develop new methods that overcome these problems. The primary benefit of this research is two-fold. First, understanding the causes of sequence reconstruction errors will help researchers better evaluate the reliability of their results, so that erroneous sequences are not inadvertently used in costly functional studies. Second, the development of more reliable methods will allow researchers to study the functions of ancient molecules that could not otherwise be examined in the laboratory. By improving the methods used to infer ancestral molecular sequences, the results of this research are expected to improve our understanding of molecular evolutionary processes and empower studies using these approaches to develop useful biological molecules. Broader impacts activities include developing new courses to train the next generation of scientists in ancestral sequence reconstruction and its uses.Developing robust ASR methods depends on understanding when and how ASR can fail. Using analyses of simulated and biological sequences, this project will test the hypothesis that ancestral sequence reconstruction is sensitive to alignment ambiguity and will develop new methods to incorporate alignment ambiguity into the ASR process. High-throughput ASR depends on accurate prediction of molecular function. This project will test the hypothesis that structure-based function prediction is sensitive to uncertainty in the underlying structural model, but accuracy can be improved by optimizing structures. This project will integrate ASR methods with structure-based function prediction to characterize the large-scale evolution of double-stranded RNA binding domains, which mediate protein-RNA and protein-protein interactions across a variety of gene contexts throughout all domains of life. By determining the links between protein sequence substitutions, structural changes and functional evolution across large protein families, this work has the potential to transform our understanding of the general principles by which molecular function evolves, fundamentally advancing molecular-evolutionary theory and providing a foundation for rational protein engineering.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.