As president and CEO of closely held Cyclica, Jason Mitakidis first became interested in computer simulation or, in silico, drug discovery while he was working as a bioinformatician at the Barcode of Life Project in 2006 and 2007. Prior to co-founding Cyclica in 2011, he gained several years of research experience in the fields of molecular oncology and structural biology, including the utilization of novel structure-based drug design methods while he was at the University Health Network and the investigation of interactions of structurally disordered protein states involved in neurological diseases while he was at the Hospital for Sick Children in Toronto. His resumé also includes a stint at JPMorgan Chase, managing a portfolio of North American corporate client accounts at the bank’s Canadian Treasury and Securities division. In this interview with BioTuesdays.com, Mr. Mitakidis discusses Cyclica’s Big Data approach to innovation and R&D productivity, including the accurate prediction of pharma side effects, the design of new drugs and the significant reduction of the time and money required to bring drugs to market.
Can you give us a brief description of Cyclica?
We are a pioneer in the development of next generation in silico discovery and testing technologies that leverage integrated biological data to improve pharma R&D productivity. We are commercializing our Ligand Express technology to address the rising cost of R&D in drug discovery and low regulatory approval rates of pre-market pharmaceutical and biopharmaceutical products. Using a cloud computing approach, our Ligand Express platform can reduce the failure risk of drug discovery projects by allowing clients to anticipate a drug candidate’s side effects prior to clinical trials, thereby enabling more informed R&D investment decisions.
How does the technology work?
Ligand Express comprises a novel drug testing approach that uses pre-computed, multi-scale structural proteomics databases that are biologically representative of conditions in the human body at 37-degrees Celsius. The platform uses these specialty databases with a chemical systems biology infrastructure to identify reliably problematic off-target protein-drug interactions that result in side effects and toxicity, which are major causes of attrition in discovery pipelines. Ligand Express is designed to be fast, accurate and also produce valuable data that medicinal chemists can use for an informed optimization process that we call Intelligent Molecular Redesign.
As a Big Data company, how do you differentiate yourself from the competition?
We differentiate ourselves in three ways. Firstly, we have generated pre-computed proteomics databases with IBM’s Blue Gene, one of the most powerful supercomputers in the world, and our own computing resources that overcome the problem of static structures. Therefore, we are able to identify a much higher percentage of protein-drug interactions than currently available technologies can. Secondly, we have developed a kinase-specific platform that provides premier, data-rich analysis on kinase drug off-target profiles, thereby allowing us to de-risk client projects and provide additional value. And lastly, our Big Data approach focuses on broad integration and indexing, unlike any technology platforms that currently exist in the biotech and pharma spaces. Similarly to how Google has indexed the information on the web for rapid search and retrieval, we have begun indexing vast stores of biological data from academic and private data for rapid search and retrieval.
In general, the in silico drug discovery and pharma sectors have not yet established clear Big Data strategies to drive innovation. There is a lot of interest in the role Big Data can play, but little innovation has occurred and the focus has been primarily on next-generation sequencing and analytics. Big Data and analytics alone won’t solve declining R&D productivity but must be integrated and actionable too.
Can you describe some cases studies you’ve done?
Pfizer’s torcetrapib is one of the most impressive case studies for the application of our technology. It was a cholesterol drug that Pfizer discontinued during Phase 3 clinical testing because of serious safety concerns, including hypertension and risk of mortality. And yet, torcetrapib had earlier passed all of Pfizer’s tests and toxicity models. Analysis using our database identified previously unknown molecular causes for the side effects.
Had you been working with Pfizer, what would have been different?
Had Cyclica’s databases been used to complement the existing data that Pfizer had on the drug, the safety risks would have been evident prior to clinical trials, with the potential of saving the company hundreds of millions of dollars. Furthermore, the data generated from Ligand Express could have been used at the late-discovery and preclinical stages to design an optimization approach that could have provided the greatest chance of minimizing the risk of side effects while still maintaining the drug’s efficacy.
How does your business model work?
The current business model is based on a managed research service. Clients include small, mid and large cap pharma and research institutes and hospitals. The client provides us with compounds for analysis, and we make the output available for secure download. We are transitioning to a product-based company with licensing and software-as-a-service revenue models.
Can you describe your partnership with IBM and its goals?
Cyclica has been granted access to IBM Blue Gene for one year, renewable for a second year, with market value of equivalent computing time equal to approximately $7-million. We have also formed a resource sharing partnership to develop new technology that is expected to optimize drug discovery and development for rare diseases. The project is called “Multi-scale characterization of protein-drug interaction networks for the FDA’s Rare Disease Repurposing Database (RDRD) using chemical systems biology and molecular dynamics methodologies.” Such multi-scale studies have never been conducted using the proprietary methods employed by Cyclica and are expected to identify drug-repurposing opportunities that other approaches would miss.
Intellectual property that arise during the course of the project will be jointly owned by the parties who created them and brought to market through, among other potential technologies, our Ligand Express platform. Other technologies that we develop may include novel methodologies, software, databases, and compounds, either in their original or redesigned form, that are found to have alternative therapeutic benefits. Repurposed compounds with a high therapeutic potential that are identified and characterized in the study will be patented and then licensed for development by pharmaceutical companies.
Can you tell us about the repurposing work you’ve done?
Sure. We ran certain FDA-approved drugs through our proteomics databases to identify putative protein-drug interactions other than with the disease target the drug was intended for. For example, with Takeda’s Actos diabetes medicine, one of the most highly ranked off-target interactions was with a protein, which we haven’t disclosed, and is known to be involved in pathogenesis of Alzheimer’s disease (AD). We were able to identify a possible mechanism of action for Actos in treating AD. When cross-referencing this result with the research literature, we found that randomized clinical trials had been performed to treat AD patients with Actos and that a reversal of cognitive decline was observed.
What’s the status of the AD work?
We have been experimenting with some modifications of Actos, seeking to improve the inhibition mechanism and have filed for a patent. In addition, we are exploring funding opportunities to perform additional research to determine if it would make sense to pursue commercialization.
What other partnerships can you tell us about?
We have an academic partnership with the Yale Center for Molecular Discovery that involves an opportunity to provide in silico support to drug discovery researchers at Yale University referred to us by the Center. The long-term objective is to approach pharma as clients together.
We also have a sales channel partnership with Sathguru Management Consultants to access emerging pharmaceutical clients in the Asian market, specifically India, Korea and China.
Do you have an exit strategy for Cyclica?
Our exit strategy will depend on whether the existing asset base of proprietary databases have the potential to generate new products and services in the future, thus creating significant value for shareholders. If we decide that’s the case, the likely exit may be an initial public offering in three-to-five years. Otherwise, the likely exit is via an acquisition by large pharma or high-tech company in two-to-three years. We have also been establishing relationships with firms that specialize in IP acquisition.
