(7ax) Stochasticity, Complexity, and Multiscale Dynamics in Cancer Progression and Drug Response

Biological systems differ from human engineered systems in three primary respects: 1) they are highly stochastic; 2) they are extraordinarily complex; and 3) they exhibit dynamics over a wide range of temporal and spatial scales. These characteristics are especially prevalent in cancer, which is a complex, adaptive disease characterized by genetic mutation, phenotypic plasticity, and cell-cell interactions among tumor cells and the tumor microenvironment. Tumors are thus highly heterogeneous and exhibit complex responses to drugs and other therapeutic interventions. Patient outcomes to therapy can vary widely, but it is not uncommon for an initial favorable response to be followed by tumor recurrence and metastasis, which is very often fatal. The mechanisms that govern drug evasion and tumor reestablishment remain poorly understood.

It is becoming increasingly clear that in order to unravel the complexities of cancer a systems engineering approach is necessary. Cancer can be seen as a reverse engineering problem, i.e., it is fundamentally a result of the breakdown of the cell’s natural tumor suppressing machinery. Reestablishing or rewiring this apparatus thus requires a deep understanding of the mechanisms that underlie normal and cancer cell function. Computational models provide a means by which current understanding of biological mechanisms can be formalized mathematically. In silico predictions can be made, tested experimentally, and the experimental results used to update and refine the model. This iterative approach is a powerful tool for obtaining a detailed and comprehensive understanding of a system over time.

Here, I present results of combined computational and experimental investigations of in vitro responses of isoclonal non-small cell lung cancer (NSCLC) and melanoma cell populations to targeted drugs. Our results point to a “bet hedging” strategy employed by cancer cell populations whereby cells diversify across multiple phenotypes in the absence of drug to increase the chances of population survival to initial drug onslaught. We hypothesize that the surviving subpopulation can persist for an extended period of time and is susceptible to genetic mutation, acting as a reservoir from which tumor recurrence may emerge. However, the subpopulation may also represent a phenotypic bottleneck that may be vulnerable to a secondary treatment. Characterizing the molecular nature of the surviving subpopulation is thus of critical importance and work is underway towards this end.

Teaching Interests:

My expertise lies in chemical kinetics, thermodynamics, and numerical methods and I also have experience in biochemistry, statistics, theoretical computer science, and software engineering. During my years as a postdoctoral researcher I have had the pleasure of mentoring numerous undergraduate and graduate students and have given numerous lectures on systems biology. I enjoy teaching and make a concerted effort to engage students by providing analogies that convey complex concepts in an easy-to-understand way. My basic philosophy of teaching is that nothing is ever as complicated as it seems and that the best way to learn something is to do it yourself. I, therefore, guide students in the right direction but ultimately leave it to them to solve a problem. It is a great feeling to witness a student having a moment of true understanding. I am prepared and look forward to teaching a full semester course at either the undergraduate or graduate level in any of the chemical engineering core disciplines.

Select Publications:

L.A. Harris*, M.S. Nobile*, J.C. Pino*, A.L.R. Lubbock, D. Besozzi, G. Mauri, P. Cazzaniga and C.F. Lopez, “GPU-powered model analysis with PySB/cupSODA,” Bioinformatics (in press). (*equal authors)

L.A. Harris*, P.L. Frick*, S.P. Garbett, K.N. Hardeman, B.B. Paudel, C.F. Lopez, V. Quaranta and D.R. Tyson, “An unbiased metric of antiproliferative drug effect in vitro,” Nat. Methods 13, 497-500 (2016). (*equal authors)

L.A. Harris, J.S. Hogg, J.J. Tapia, J.A.P. Sekar, S. Gupta, I. Korsunsky, A. Arora, D. Barua, R.P. Sheehan and J.R. Faeder, “BioNetGen 2.2: Advances in rule-based modeling,” Bioinformatics 32, 3366-3368 (2016).

J.S. Hogg*, L.A. Harris*, L.J. Stover, N.S. Nair and J.R. Faeder, “Exact hybrid particle/population simulation of rule-based models of biochemical systems,” PLoS Comput. Biol. 10, e1003544 (2014). (*equal authors)

L.A. Harris*, J.S. Hogg* and J.R. Faeder, “Compartmental rule-based modeling of biochemical systems,” Proceedings of the 2009 Winter Simulation Conference, M.D. Rossetti, R.R. Hill, B. Johansson, A. Dunkin, and R.G. Ingalls, eds., pp. 908-919 (2009). (*equal authors)

L.A. Harris and P. Clancy, “A ‘partitioned leaping’ approach for multiscale modeling of chemical reaction dynamics,” J. Chem. Phys. 125, 144107 (2006).

L.A. Harris and A.A. Quong, “Molecular chemisorption as the theoretically preferred pathway for water adsorption on ideal rutile TiO2(110),” Phys. Rev. Lett. 93, 086105 (2004).