Person:

Bachman, John

Mathematical equations are fundamental to modeling biological networks, but as networks get large and revisions frequent, it becomes difficult to manage equations directly or to combine previously developed models. Multiple simultaneous efforts to create graphical standards, rule-based languages, and integrated software workbenches aim to simplify biological modeling but none fully meets the need for transparent, extensible, and reusable models. In this paper we describe PySB, an approach in which models are not only created using programs, they are programs. PySB draws on programmatic modeling concepts from little b and ProMot, the rule-based languages BioNetGen and Kappa and the growing library of Python numerical tools. Central to PySB is a library of macros encoding familiar biochemical actions such as binding, catalysis, and polymerization, making it possible to use a high-level, action-oriented vocabulary to construct detailed models. As Python programs, PySB models leverage tools and practices from the open-source software community, substantially advancing our ability to distribute and manage the work of testing biochemical hypotheses. We illustrate these ideas using new and previously published models of apoptosis.

Mitochondrial outer membrane permeabilization is a key step in the apoptotic cell death program, regulating life-death decisions in response to cytotoxic drugs and other forms of cell stress. In this thesis I use mathematical modeling of a reconstituted biochemical system to identify and integrate mechanisms of apoptotic pore formation. A key bottleneck in using mathematical models to characterize mechanisms has been the difficulty of efficiently creating and revising alternative models and evaluating them against data. This problem is addressed through the use of a software framework, PySB, that allows ensembles of models to be transparently described using tools and approaches from computer programming. These alternative hypotheses can then be evaluated against data using methods from Bayesian statistics for discrimination of models with varying numbers of (possibly non-identifiable) parameters. Using this framework, calibration of a set of models to in vitro kinetic measurements of the membrane insertion of Bax identifies a conformational intermediate associated with BH3-only:Bax complex formation and membrane association but not pore formation. Functional measurements of Bax point mutants from tumors show that the pore formation process can be blocked at the transition into or out of this intermediate, preventing mitochondrial permeabilization. In another study, model-based analysis of Bax insertion and permeabilization kinetics across a range of BH3-only, Bax, and liposome concentrations reveals the context-dependence of the mechanisms regulating pore formation. Bax recruitment is shown to depend on liposome concentration kinetically but not stoichiometrically, whereas cBid recruitment is shown to be limited at high cBid:liposome concentrations. I show that Bax distribution among liposomes is dependent on the presence of pre-existing pores, and that pores grow to include large numbers of Bax monomers but have a minimum size of four subunits. More generally, these studies serve as examples of how ensemble modeling can be used to integrate information about complex mechanisms from disparate sets of experimental observations.

Abstract Word models (natural language descriptions of molecular mechanisms) are a common currency in spoken and written communication in biomedicine but are of limited use in predicting the behavior of complex biological networks. We present an approach to building computational models directly from natural language using automated assembly. Molecular mechanisms described in simple English are read by natural language processing algorithms, converted into an intermediate representation, and assembled into executable or network models. We have implemented this approach in the Integrated Network and Dynamical Reasoning Assembler (INDRA), which draws on existing natural language processing systems as well as pathway information in Pathway Commons and other online resources. We demonstrate the use of INDRA and natural language to model three biological processes of increasing scope: (i) p53 dynamics in response to DNA damage, (ii) adaptive drug resistance in BRAF‐V600E‐mutant melanomas, and (iii) the RAS signaling pathway. The use of natural language makes the task of developing a model more efficient and it increases model transparency, thereby promoting collaboration with the broader biology community.