A signaling visualization toolkit to support rational design of combination therapies and biomarker discovery: SiViT

Abstract

Targeted cancer therapy aims to disrupt aberrant cellular signalling pathways. Biomarkers are surrogates of pathway state, but there is limited success in translating candidate biomarkers to clinical practice due to the intrinsic complexity of pathway networks. Systems biology approaches afford better understanding of complex, dynamical interactions in signalling pathways targeted by anticancer drugs. However, adoption of dynamical modelling by clinicians and biologists is impeded by model inaccessibility. Drawing on computer games technology, we present a novel visualization toolkit, SiViT, that converts systems biology models of cancer cell signalling into interactive simulations that can be used without specialist computational expertise. SiViT allows clinicians and biologists to directly introduce for example loss of function mutations and specific inhibitors. SiViT animates the effects of these introductions on pathway dynamics, suggesting further experiments and assessing candidate biomarker effectiveness. In a systems biology model of Her2 signalling we experimentally validated predictions using SiViT, revealing the dynamics of biomarkers of drug resistance and highlighting the role of pathway crosstalk. No model is ever complete: the iteration of real data and simulation facilitates continued evolution of more accurate, useful models. SiViT will make accessible libraries of models to support preclinical research, combinatorial strategy design and biomarker discovery.

Dense agent-based HPC simulation of cell physics and signaling with real-time user interactions

“…In the context of drug discovery, an interactive simulation is particularly useful when exploring the emergent behaviors resulting from complex interactions between agents. For example, combination therapies use the effects of multiple drugs taken at different times; the effects of these upon the cells can be more effectively seen in an interactive simulation (Bown et al, 2017).…”

“…SiViT (Bown et al, 2017) simulates and visualizes the response of an ovarian cancer cell to different drug interventions. The existing implementation of SiViT operates in three stages: first, the user defines a treatment regime, then the simulation runs, and finally, the simulation results are presented graphically.…”

Narratives for Drug Design

“…This is, in part, because the single therapies are acting in the context of a range of mechanisms that compensate for and adapt to perturbations (here, drug action): these mechanisms include cross-talk, feedback loops, differential sensitivities to change across the network and changes in network structure in response to drug action. This means that targeted therapies can impact beyond their point of application, and often in ways that are difficult to anticipate (Bown et al 2016). These features then limit efficacy of any single therapy, and patient resistance to a drug is a key challenge in anti-cancer therapy design.…”

“…We propose that our technology is a vehicle to support reader-driven narratives, but is not in of itself a narrative. This technology, SiViT (Bown et al 2016), turns a complex model into an interactive animation, allowing the cancer specialist intuitive access to complex systems models otherwise inaccessible. SiViT is able to represent graphically the network structure of models of cell signalling, such as that described in Figure 1.…”

“…a signaling network visualisation with a pop-up dialogue box (Inset 1, bottom right) for amending drug regime and mutational status together with an inset with magnified detail (Inset 2, bottom left) taken fromBown et al (2016). See http://www.impactjournals.com/oncotarget/index.php?journal=oncotarget&page= article&op=view&path%5B%5D=8747 for figures and related movies.…”