Cracking the viral code
The havoc and tragedy wrought by the coronavirus pandemic remind us of the perils of an unchecked virus infecting the world. One reason why it’s difficult to quickly design and manufacture effective antiviral treatments, especially for novel viruses, is the lack of a systems-level understanding of the viral mechanisms that damage host cells. We need to combine experimental data with computer simulations to create virtual versions of real biological systems in order to analyze their systems-level dynamics and develop targeted therapeutic antidotes.
Even relatively simple viral genomes generate complex infections that are difficult to disrupt. For example, the Ebola virus has only seven genes, yet it can spread rapidly to organs throughout the body and cause tissue damage, diarrhea, hemorrhage, and death. My lab is investigating this virus as a model for a systems approach to unraveling the deeply-connected, multi-scale dynamics that sustain the robust replication and spread of Ebola and other viruses.
Our central hypothesis is that the complexity and virulence of viral infections can be laid at the doorstep of a series of interconnected subcellular and intercellular interactions, and that throwing a wrench into these dynamic processes can lead to novel antiviral strategies. To identify these interactions, we’re developing computational models across multiple time and spatial scales, designing new methods of analysis, and testing our model predictions experimentally by using a mini-genome replica of the Ebola infection.
After attaching to and entering a host cell, several Ebola virus proteins engineer the primary transcription (generating mRNA molecules that can be translated into new viral proteins), as well as genome replication necessary to produce more functional viruses. A series of complex and highly interrelated events must take place with precisely the right magnitude and timing to create a sustained production cycle of virions (the infective form of the virus). While individual protein functions have been identified, the effects of these individual mechanisms in larger systems, and how the interconnected mechanisms collectively drive multi-scale dynamics, remain largely unknown.
We want to lift that veil by building progressively more complex models closely coupled with experimentation, pushing the boundaries of computational models in biological research. Individual proteins impact cells and tissues via complex interacting networks, but they are largely studied in isolation in experiments. Mathematical and computational models can integrate multiple interactions and experimental datasets to uncover previously unidentifiable patterns.
We’re using advanced computational and algorithmic tools to power our research. For example, to quantify uncertainty in our predictions, we use a Bayesian Markov Chain Monte Carlo algorithm to sample probability distributions of model parameters. We will use a stochastic (probabilistic), rule-based model, and solve systems via the Gillespie algorithm.
The great thing about the simulations is that, because they are just equations and computer code, we can measure any output any time, or remove any individual mechanism to test the overall impact on the system. Computational models provide more flexibility that we can use to inform the experiments and help evolve our understanding of this complex system.
There’s also an educational component to our work. Biomedical engineering operates at the intersection of physical and life sciences, raising both challenges and opportunities for education. We not only want to develop computational tools and techniques; we also want to train the scientists who use them to advance computationally-driven biological discovery.
Because our work straddles computation and biology, we want to increase the exposure of biology students to computational approaches at the high school and undergraduate levels. We aim to do this in a way that will enable instructors and educators to easily pick up and use our tools/modules and integrate them into their classes. We also want to develop the interdisciplinary skill set in engineering students through close collaboration with biological experts and peers. Our hope is that the interdisciplinary research, together with interdisciplinary educational efforts, will help build strong and lasting connections among future biologists and engineers.
This interlinked research and educational work is sponsored by a National Science Foundation (NSF) Early Career Development (CAREER) award. We’re collaborating with Robert Stahelin, Retter Professor of Pharmacy and Professor of Medicinal Chemistry and Molecular Pharmacology at Purdue University; we will use experimental data from his lab to estimate parameters for our simulations and test our predictions. The research was initiated with funding from the Indiana Clinical and Translational Sciences Institute (CTSI) Project Development Teams.
Overall, we need novel, quantitative and systems-level understanding of how viral mechanisms interact to facilitate viral propagation in order to advance virus research. On the educational side, we need to blend engineering and biology training to support our young scientists in these interdisciplinary fields. This combination of research and education can help develop a new model to integrate computational methods into biology, equipping the next generation of scientists to blaze new trails into truly interdisciplinary science.
Weldon School of Biomedical Engineering
College of Engineering