follow us us on Twitter
Subscribe to our mailing list
Creating a new paradigm of drug developmentNov 25, 2019
by David Kita
We founded Verseon back in 2002 because we noticed that, despite billions of dollars being invested in pharmaceutical R&D, new small-molecule drug approvals had stagnated. The prevailing mechanisms for drug development at the time, and still to this day, are high-throughput screening and follow-on drug design. These conventional methods are long overdue for an overhaul.
Many diseases are caused by a misfunction of certain proteins in the body. The challenge of developing new small-molecule medicines is finding molecules that bind to these disease-causing proteins to restore or limit their function. High-throughput screening tries to tackle this problem using trial and error: a number of previously synthesized molecules are tested against the disease-causing protein in the lab. In this process, only those molecules that have previously been produced can be tested, which creates a bottleneck to the discovery of novel chemical matter. The follow-on design approach modifies an existing drug, yielding drug candidates that closely mimic the pharmacology of approved drugs. Again, completely new small-molecule drugs are rare.
With Verseon, we set out to create a new paradigm for finding novel drugs. We wanted to establish a more efficient way to identify potent binders to target disease-causing proteins using computational modeling. With this goal in mind, we started to develop our molecular modeling engine.
In the past, computational methods in drug discovery have typically approached protein-drug interactions with heuristic models, and later progressed to training-set-based methods, eventually turning to machine learning. But those approaches suffer from limited accuracy and rarely generate truly novel drugs. We decided to attack this problem the way it should be: as a physics problem. In fact, it’s a fiendishly challenging physics problem with aspects of electrostatics, quantum mechanics, molecular dynamics, and statistical physics. With that in mind, we assembled a team of physicists and computational scientists to develop a drug discovery platform capable of accurately predicting whether a small molecule will bind to a target protein in solution.
Along the way, we have developed advanced physics-based models that faithfully capture the underlying molecular interactions between the target protein and small molecule. We have also developed sophisticated optimization algorithms that can robustly accommodate the many degrees of freedom of molecular complexes, including protein flexibility and water mobility.
Instead of licensing out the resulting disease-agnostic computational platform, we decided to use it to drive our own drug discovery programs. Because the initial phases of our process are computer-driven, we are not limited to pursuing only one program in-house, but we run multiple programs in parallel. Then, as drug candidates are nominated for progression toward the clinic, computational resources are freed up for the exploration of other drug programs.
This is a constant cycle of discovery with the potential to explore many disease indications. The ability to address just about any disease with known 3D structure of the target protein has since allowed us to develop a diverse pipeline. Verseon’s strategy diverges from the industry norm and that will continue to set us apart as we navigate our drug candidates through clinical trials.
David Kita is co-founder and Vice President of R&D at Verseon.
- drug discovery -