Re-Inventing Drug Discovery

Denovicon uses an innovative combination of machine learning and physics-based modeling to streamline the development of small molecule therapeutics, dramatically reducing R&D timelines and improving clinical success.

Breaking Through the Challenges of Drug Discovery

The pathway to a marketed drug is fraught with numerous challenges. Statistically, only about 14% of compounds that enter clinical trials succeed, with even lower success rates in oncology, sometimes dropping to 3%. Most drug-discovery chemistry occurs within a commonly explored space with the more challenging compounds rarely being made. A drug discovery strategy that uses a combination of physics-based, machine learning/AI, and generative design methods could dramatically improve the overall success.

Denovicon employs a hybrid approach that is not biased towards any specific computational technique, but rather focuses on utility and durable integration. BIOVIA’s data science and drug discovery solutions play a significant role by allowing various molecular data processes to be streamlined into a seamless workflow. In this way, Denovicon reduces the R&D timeline from the typical 5–7 years to 1–2 (and in some cases under a year), with the potential to improve clinical success upwards of 40-50% (versus the current rate of 14%).                             

A Hybrid Platform to Enhance Efficiency  

Denovicon’s hybrid approach involves a sophisticated machine learning–physics-based workflow to sift through billions of compounds to identify the most viable leads within a desirable chemical space. In its kickoff project on next-generation PARP1 inhibitors, this approach resulted in an unprecedented 80% success rate (using a high bar of 50% inhibition @ 1µM) as opposed to the traditional HTS hit rate, which ranges from 0.5–1%, but is usually < 0.1%. While only 236 compounds were tested in total, 81% had an IC50 ≤ 100 nM and 47% had an IC50 ≤ 10 nM with the best compound already achieving a druglike affinity of < 1 nM.

Denovicon’s platform goes beyond lead identification, addressing properties such as solubility, cell permeability, oral bioavailability, ADME/PK, toxicity, and many others required to take a lead through optimization to clinical candidate.  Denovicon’s platform is particularly enhanced by its use of quantum machine learning to dramatically lower the computational expense of traditional quantum mechanics calculations. These surrogate models allow Denovicon to achieve high accuracy at scale while lowering the computational burden.

 Streamlining Processes with BIOVIA

BIOVIA’s data science and drug discovery solutions synergize with Denovicon’s platform, offering seamless integration and unification of various computational approaches. In particular, Denovicon uses BIOVIA’s Pipeline Pilot to integrate third-party software, open-source tools, and its proprietary code, which creates a flexible and powerful environment for designing drug discovery workflows that are easily reproducible, reusable, and upgradable.

“We’re not just finding hits, we’re finding high-quality leads that are quickly turned into clinical candidates, in a fraction of the time of traditional drug discovery.” – Scott Bembenek, CEO and Founder, Denovicon Therapeutics      

Achieving Success in Drug Development

Denovicon’s innovative hybrid platform has led to significant advancements in developing unique, potent inhibitors in the PARP and DNA Damaged Response Pathways (DDR) spaces. As with the PARP1 project, Denovicon has also identified potent and high-quality leads for its more recent PARP2 and PARP7 projects. The PARP7 project, which is a collaboration with BIOVIA, further showcases the platform’s lead-identification capabilities where again an unprecedented success rate of 59% was achieved.

We consistently find double-digit – even single-digit nanomolar leads – allowing us to immediately jump into SAR development and optimization towards a clinical candidate without ever getting stuck in the classic hit-to-lead bottleneck.

– Scott Bembenek, CEO and Founder, Denovicon Therapeutics  

Strategic Collaborations and Future Directions

Denovicon leverages its drug-discovery expertise, its vast life-sciences network, a strong Scientific Advisory Board, and strategic collaborations (with partners like Dassault Systèmes and BIOVIA) to maintain a competitive edge. This along with its approach of outsourcing chemistry and early-stage biology while partnering for later-stage disease models allows for efficient operation without a large in-house footprint.

In summary, Denovicon employs an innovative approach to drug discovery, powered by AI, quantum machine learning, physics-based modeling, as well as strategic use of BIOVIA solutions. The company’s focus on reducing timelines, improving success rates, and developing high-quality compounds marks a significant shift in how pharmaceutical research is currently being conducted.

Our strategy is about developing and leveraging the best tools and collaborations – like the ones we have with BIOVIA and Dassault Systèmes (especially through the 3DEXPERIENCE Lab)  – to accelerate drug discovery and deliver effective therapies to patients faster.

– Scott Bembenek, CEO and Founder, Denovicon Therapeutics

---

📩Want to find out the latest news about BIOVIA events, customer stories, blogs and more? Join the monthly newsletter today!