Constructing the Simulated Cell
- Turbine’s Simulated Cell is a directed and weighted signaling network, which incorporates 25 important signaling pathways acting in cancerous cells, including MAPK, ErbB, Notch and Hedgehog, describing all hallmarks of cancer.
- Turbine’s Simulated Cell is 10x the size of the largest published signaling network usable for dynamic simulations. It currently incorporates 1100+ proteins and other second messenger molecules and their over 3000 interactions. The network incorporates 25 important signaling pathways acting in cancerous cells, including MAPK, ErbB, Notch and Hedgehog, describing all hallmarks of cancer. It can be customized for any other complex disease as well.
- The network itself is a directed and weighted signaling network, where the vast majority of the links represent direct protein interactions and nodes represent unique protein isoforms.
- Every node and interaction undergoes rigorous manual curation. The functional behavior of the Simulated Cell was tested on dozens of well-described mechanisms, which the cell displayed without error.
Customizing the Simulated Cell to cancer type and patient
- Turbine’s simulations closely resemble actual in vitro and in vivo interactions because we use available omics data to customize our cell model to a cell line or patient sample. Genetic mutations and transcriptomic information are both incorporated.
- The genomic and transcriptomic characteristics of cancer cells vary widely not just among patients, but often inside the same tumor as well. Turbine uses available OMICS data to customize the cell model to a cell line or patient sample, enabling simulations to closely mimic the actual, in vitro or in vivo interactions.
- Genetic mutations, either congenital or acquired, play a key role both in the development and the propagation of cancer. Both gain and loss of function mutations and their unique characteristics are taken into account.
- The cell model can also incorporate transcriptomic information from any given human tissue or cell line to make predictions even more accurate. The given sample’s expression profiles are compared to transcriptomic datasets from a wide variety of healthy tissues, to find out the relative difference between them, which is then applied to the Simulated Cell.
- After completing the patient or cell line specific cancer cell model, we can either construct similar control models for healthy tissue samples from the patient or use previously available in house developed in silico healthy cell models as controls. We use these to model the potential toxicity of any therapy or therapy combination, minimizing additive toxicity.
Compiling the simulation’s drug library
- To find effective treatment combinations, available drugs and their targets are mapped to the simulated cell’s nodes. Compounds can include a client’s entire compound library, approved treatments for any disease, or other available public or in-house molecules. No structural data is required.
- To find effective treatment combinations, the available drugs and their targets, with relative affinity (if available) are mapped to the Simulated Cell’s nodes.
- A drug library can be compiled from a client’s entire compound library, approved treatments for any disease, or from other available public or in-house sources. Turbine doesn’t require the structure of the given compound, only the protein target profiles and relative affinity of each protein to the compound based on earlier bioactivity data from experiments or chemoinformatic modeling.
- It’s important to note that Turbine’s artificial intelligence is not limited to already available treatments or molecules, and can identify novel drug targets or target combinations if simulations show that new compounds targeting these could boost tumor response.
Modeling therapy response with the Turbine simulator
- Turbine’s network simulator runs thousands of simulations on the customized cell to model the behavior of the network before and after therapies are administered. Turbine looks for so-called attractors, the steadiest states of the network that define how a cell functions or reacts to certain interventions.
- The customized cancer cell model is brought to life by Turbine’s dynamic network simulator. The simulator models the behavior of the network before and after therapies are administered, biologically accurately representing how signals propagate in the cell’s protein network, and how the cell ultimately responds to the intervention – either initiating apoptosis, halting the cell cyle or proliferating further. Turbine’s simulator has been developed to incorporate interactions resulting in both translational and post-translational modifications.
- This allows running thousands of simulations for each therapy and their combinations, and looking for states that the network enters most often. Complex networks like human cells can have billions of states, but these so-called attractors are the steadiest ones that define how a cell functions or reacts to certain interventions. To accommodate dosage response, Turbine simulates therapy regimes in various doses, effectively mapping the dose-response curves of each compound.
- The simulator’s results have been compared to published research and both in vitro and in vivo measurements to be biologically accurate, correctly predicting drug response for 40% of drugs compared to in vitro measurements.
Designing optimal therapies with artificial intelligence
- As simulations are very fast, Turbine uses an artificial intelligence to select and discover the most effective therapy combinations among trillions of options. Our AI is a global optimization algorithm developed specifically to find effective combinations in the large search spaces of dynamic networks.
- Executing a single simulation is lightning fast, thanks to our dynamic simulator that is 1000x faster than the best published alternative. At this speed, we can use an artificial intelligence to guide the simulations to select and discover the most effective therapy combinations among untold trillions of possibilities, compressing years of unguided, brute force simulation time into weeks.
- The Turbine AI is a global optimization algorithm refined especially for the purpose of finding combinations in the large search spaces of dynamic networks, supplemented by years of research in the theory of dynamic biological systems. The Turbine AI starts by creating huge, complex combinations, then filters them down to their most essential, effective components - which are applicable in a clinical or experimental setting.