High-Performance Parallel Simulations for Whole-Cell Modeling
Ali Navid | 19-ERD-030
Whole cell models (WCMs) are a new multi-scale class of cellular mathematical models that simulate all known mechanistic processes in an organism. These powerful predictive biology tools incorporate a diverse array of chemical pathways, cellular phenomena, and computational methods. WCMs are powerful predictive biology tools that permit examination of many complex biosystem properties and how interactions between various bioprocesses affect system phenotypes. The mechanistic understanding gained from these simulations can be used to obtain new insights that can be used for important applications such as engineering a biosystem's behavior to achieve desired outcomes such as production of compounds of interests like drugs or biofuels. Efficient, and rapid simulation of WCM has been stymied by nonexistence of an efficient parallelizable simulator that can run the models using high performance computing (HPC) resources.
In this project we developed several tools and techniques to overcome this challenge. We developed a general purpose high-performance, multi-scale, parallel whole cell simulator (WCS) that can be used to execute several critical modeling algorithms that are used in system-level biological simulations. To achieve this goal, we have developed a method to ingest systems biology markup language (SBML)-based models as input into WCS, a method for analyzing the organizational characteristics of the models in order to restructure the network and expose dependencies, all while parallelizing the network, a parallel implementation of a stochastic simulation algorithm, a representative workload of a large biochemical reaction network in order to examine the efficiency of our WCS, and an algorithm to allow for seamless combination of stochastic simulations with deterministic ordinary differential equations (ODE)-based calculations. Our efforts have greatly shortened the time needed to simulate complex WCMs and we are in the final stages of using our WCS to simulate a sizeable pan cancer system model for purpose of examining an important oncogenic process.
Computational tools capable of accurately predicting the transient behavior of biosystems will revolutionize systems biology and accelerate design of new experiments. They will improve our ability to implement bioengineering projects, identify new therapeutic targets, and prevent occurrence of adverse drug reactions. This project contributed to the chemical and biological security mission focus of Lawrence Livermore National Laboratory (LLNL) and falls directly in the R&D area "Simulating Complex Biological Systems" and the overall Biological Applications of Advanced Strategic Computing (BAASIC) initiative. The models that can be simulated using our WCS can examine in unprecedented detail mechanisms of critical bioprocesses such as host/pathogen interactions, biosynthetic pathways, identification, formation, and life cycle of disease biomarkers. The project has leveraged LLNL's HPC capabilities, as well as expertise in parallel discrete events modeling/analysis, big data sciences, and computational biology to develop the most advanced multi-math simulator of system-level biological models. This LDRD has place LLNL at the forefront of the WCM simulation field. Lab investigators have begun developing new models and designing new experiments that will rely on WCM predictions. The new WCS is already resulting in new collaborations with industry, academia, as well as other federal agencies involved in biological research such as NIH, DOD and DHS.
Publications, Presentations, and Patents
Konstantia Georgouli, Jae-Seung Yeom, Robert Blake and Ali Navid. 2021. "Import and parameterization of system-level models for assessing the performance of a whole-cell model simulator." 14th Great Lakes Bioinformatics (GLBIO) conference, poster presentation. May 10-13, 2021.
Yeom, Jae-Seung, Konstantia Georgouli, Robert Blake, and Ali Navid. 2021. "Towards Dynamic Simulation of a Whole Cell Model." In Proceedings of the 12th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, August 2021 Article No.: 82, https://doi.org/10.1145/3459930.3471161.
Cemal Erdem, Arnab Mutsuddy, Ethan M. Bensman, William B. Dodd, Michael M. Saint-Antoine, Mehdi Bouhaddou, Robert C. Blake, Sean M. Gross, Laura M. Heiser, F. Alex Feltus, and Marc R. Birtwistle. 2021. "A Scalable, Open-Source Implementation of a Large-Scale Mechanistic Model for Single Cell Proliferation and Death Signaling." Nature Communications, accepted.