Find biomarkers of sensitivity and drug resistance
Turbine's AI combinines OMICS profiles, signaling activity, and simulated drug response of different cancer cells to find complex biomarkers of sensitivity or resistance, such as mutational or gene expression patterns.
Before running in silico experiments, Turbine’s Simulated Cell is customised with tumor specific OMICS data. By combining the OMICS profiles, signaling activity, and simulated drug response of different cancer cells, Turbine’s AI can find complex biomarkers of sensitivity or resistance, such as mutational or gene expression patterns.
Such biomarkers can then be used to select patients who will respond particularly well to a drug or combination. The result is a faster approval process, potentially adding years to a drug’s time on market before patent protection ends.
Identify biomarkers of sensitivity and resistance
Based on OMICS data, Turbine can identify biomarkers that mark mutations with increased sensitivity or resistance to treatment.
Simulating therapy effects on healthy cells and different mutational profiles can allow Turbine to predict toxicity.
Compare responder and non-responder patient populations
Turbine can work its way back from the phenotypic behavior of patient cells or cell lines to identify distinctive molecular patterns or biomarkers that differentiate those who respond from those who do not.
Select best fitting patients
After identifying biomarkers, Turbine can help trial teams select patients who’d benefit most from entering the trial.
“The first attempt at an integrative view of the effects of
cancer mutations, long missing in molecular biology.”
Design effective combination therapies
Given a drug candidate and planned indications, Turbine can identify the most promising combination candidates to trial. Additionally, Turbine can simulate any cell based on existing OMICS data, thus enabling the testing of combination therapies in any indication, cell line, or patient.
Designing clinical trials that prove clinical benefits clearly enough for fast tracked approval is a monumental challenge. Much of its complexity is due to the multitude of combinations possible for a targeted cancer drug candidate, and the heterogeneity of each potential trial participant’s tumor.
Turbine tackles both challenges - given a drug candidate and planned indications, it can identify the most promising combination candidates. Turbine can also simulate any cell based on existing OMICS data, thus enabling the testing of combination therapies in any indication, cell line, or patient.
Uncover synergistic effects
Combining a lead with approved drugs or other candidates, synergistic effects can be calculated.
Block escape routes of cancer
Design combinations that minimize the chance of developing resistance by blocking the most likely evolutionary escape routes.
Simulate optimal dosage
Based on simulations, Turbine estimates dose-response curves for each compound and combination, similarly to in vitro experiments.
Accurate for rare cancer strains
Turbine can customize its Simulated Cell for most cancer types, even rare strains.
8,000,000 leads screened in 1 day to identify
best treatment for indication
Extend and repurpose drug lines with novel combinations
Turbine can customize its Simulated Cell to almost any cancer indication, then let its artificial intelligence simulate and analyze the impact of over 8 million treatment options per day. Thus, Turbine can identify potential candidates for repurposing, and indications with the highest predicted response rate, to launch trials to extend drug lines faster and with more certainty.
There are approximately 45,000 2-drug combinations of approved cancer therapies, and 4,500,000 3-drug ones. This complexity is daunting, but also means that literally thousands of therapy options lie hidden in the current approved drug library.
Turbine can customize its Simulated Cell to almost any cancer indication, then let its artificial intelligence simulate and analyze the impact of over 8 million treatment options per day, quickly homing in on potential candidates for repurposing. By selecting the treatments most likely to work and indications with highest predicted response rate, trials to extend drug lines can be launched faster and with more certainty.
Repurpose failed leads
By simulating the effects of your failed candidates combined with available approved drugs, Turbine can identify opportunities to massively increase clinical benefits or find new therapeutic areas.
Find line extension opportunities
Given your drug’s target profile, Turbine can identify other approved compounds that would provide synergistic effects, opening up new indications.
High throughput therapy screening
Unlike existing in vitro methods, Turbine can simulate a compound’s effect in a fraction of a second, providing insight into the complex biological background of any response on the level of cellular signaling.
Test on any cancer type
Turbine can customize its Simulated Cell for most cancer types, even rare strains.
Every combination of FDA approved cancer drugs
tested in ~4 WEEKS
Understand the biology of cancer
Turbine's customizable Simulated Cell models intracellular phenomena in a biologically accurate way. Our AI can manipulate the cell to find the fundamental drivers of its cellular activities. Results can help in the generation of new hypotheses, and give direction to laboratory research.'
Turbine’s Simulated Cell combines several OMICS layers with detailed modeling of cellular signaling pathways. This combination yields a flexible, yet biologically accurate model of intracellular phenomena. Researchers can change molecular characteristics at will, observe resulting changes, and generate new hypotheses.
Turbine’s AI helps you understand how any cancer works by finding the fundamental drivers of its cellular activities. What’s more, it turns these findings into actionable recommendations that can be directly validated in laboratory experiments.
Insight into cellular pathway activity
Turbine realistically models intracellular molecular interactions and signaling pathway activities, and displays them on a temporal scale, ready for detailed analysis.
Understand mechanism of action
Turbine can uncover a drug's exact mechanism of action on a molecular level, providing extra understanding compared to in vitro experiments.
Block pathways of cancer evolution
Turbine can model the possible directions evolution can take in a tumor. This understanding can help predict and prevent acquired resistance or relapse.
Discover biological mechanisms
Turbine can discover the biological mechanisms leading from one state of a cell to another, such as pre and post relapse samples, and offers ways to suppress or amplify them.
As few as 5 patients’ sequenced tumor data is enough
to model even rare cancer strains.
Speed up cancer drug discovery
We arrange for a short demo of Turbine’s workflow and results, and to discuss how we can help your company or research.
If you prefer, we tap already available OMICS or in vitro data in your database to deliver novel insight and drug combinations without any additional work on your part.
Prospective collaboration can be started to discover new therapeutic potential in your drug library, analyze cellular behavior or design effective new trials.
Questions about Turbine?
Does Turbine need structural data about modeled compounds?
No, Turbine only needs to know the protein target profiles and binding affinities (if available) of the drugs we simulate.
What kind of data does Turbine need to start simulations?
We use genomic (mutation, CNV), transcriptomic (microarray or RNA-seq) and proteomic profiles (if available) from sequenced patient tumors, cell lines or xenografts; as well as the protein target profile and binding affinities of drugs.
Can Turbine utilize my existing data?
Yes, Turbine can use your existing in vitro and in vivo measurements to find novel drug combinations - no new measurements are needed to start collaboration.
What kind of model systems can Turbine utilize? (xenograft, animals, etc)
We can predict therapy response on model cell lines, generic tumor subtypes (based on several sequenced tumor samples), xenografts, primary cultures and even individual patients (based on their sequenced tumor data).
Why is Turbine better than big data based approaches?
Turbine is not a traditional "big data" approach and does not depend on large Bayes networks and thousand or million data point databases of therapy response or experimental data. Our Simulated Cell enables us to model how cells function - a much better understood area than treatment response to various drugs. Consequently, our results are much richer and flexible: we can predict biomarkers of sensitivity and resistance, identify novel drug targets, predict dose dependent synergy, design combinations of three or more drugs, and even model additive toxicity in silico.
How can you simulate millions of experiments in a realistic timeframe?
We utilize an AI to guide the simulated experiments. In this manner, running a million simulations covers many of the best hypotheses in an exponentially larger search space of trillions - never simulating experimental setups unlikely to yield significant results.
Does Turbine have any success stories so far?
We have concluded a validation study that found Turbine can accurately predict dose-response for 4 out of 10 treatments (p<0.001). This result is an order of magnitude better than traditional pharma R&D methods. To learn more, please contact us.
How do you ensure shared data is kept safe?
We use encrypted storage and very thorough access management to safeguard all client data. No personally identifiable information is used by Turbine to ensure patient privacy. Turbine runs all simulations in the cloud, which guarantees physical security.