Simulation of Biochemical Pathway Adaptability Using Evolutionary Algorithms

Bosl, W J

Lawrence Livermore National Laboratory

关键词: Lawrence Livermore National Laboratory;    Microorganisms;    Genetics;    99 General And Miscellaneous//Mathematics, Computing, And Information Science;    Genes;   

DOI  :  10.2172/917509
RP-ID  :  UCRL-TR-209542
RP-ID  :  W-7405-ENG-48
RP-ID  :  917509

美国|英语

来源: UNT Digital Library

【 摘 要 】

The systems approach to genomics seeks quantitative and predictive descriptions of cells and organisms. However, both the theoretical and experimental methods necessary for such studies still need to be developed. We are far from understanding even the simplest collective behavior of biomolecules, cells or organisms. A key aspect to all biological problems, including environmental microbiology, evolution of infectious diseases, and the adaptation of cancer cells is the evolvability of genomes. This is particularly important for Genomes to Life missions, which tend to focus on the prospect of engineering microorganisms to achieve desired goals in environmental remediation and climate change mitigation, and energy production. All of these will require quantitative tools for understanding the evolvability of organisms. Laboratory biodefense goals will need quantitative tools for predicting complicated host-pathogen interactions and finding counter-measures. In this project, we seek to develop methods to simulate how external and internal signals cause the genetic apparatus to adapt and organize to produce complex biochemical systems to achieve survival. This project is specifically directed toward building a computational methodology for simulating the adaptability of genomes. This project investigated the feasibility of using a novel quantitative approach to studying the adaptability of genomes and biochemical pathways. This effort was intended to be the preliminary part of a larger, long-term effort between key leaders in computational and systems biology at Harvard University and LLNL, with Dr. Bosl as the lead PI. Scientific goals for the long-term project include the development and testing of new hypotheses to explain the observed adaptability of yeast biochemical pathways when the myosin-II gene is deleted and the development of a novel data-driven evolutionary computation as a way to connect exploratory computational simulation with hypothesis-driven experimentation. This LDRD will focus on developing prototype software for the evolutionary computation and demonstrating its efficacy on a well-known biochemical pathway in yeast. Expected outcomes from this LDRD project included a demonstration of computational modeling of evolvability in a biochemical pathway, an important collaboration with the Systems Biology department at Harvard University, several proposals to secure external long-term funding from one or more sources and the nucleus of a new, focused research effort at LLNL in computational genomics, focused principally on Genomes to Life goals. All of these goals were achieved.