Superluminal Medicines Secures $33M Seed Funding to Accelerate AI-Driven GPCR-Targeted Drug Discovery

Recently, VCBeat learned that biotechnology company Superluminal Medicines (hereinafter referred to as “Superluminal”) announced the completion of a $33 million seed financing round. The round was led by RA Capital Management, with participation from Insight Partners, NVIDIA, and Gaingels. The funds will be used to advance its small-molecule drug development pipeline and accelerate the research and development of G protein-coupled receptor (GPCR)-targeted therapies.

As a biotech startup founded just over a year ago, what has enabled Superluminal to win the favor of investment institutions in such a short time? And how has it emerged as a rising star in the small-molecule drug market?

Superluminal, founded in 2022 and headquartered in Boston, USA, is an innovative small-molecule drug R&D company centered on AI technology. It has built a drug discovery pipeline platform by seamlessly integrating AI with biology, chemistry, and computational science to enhance the speed and accuracy of drug development.

Cony D’Cruz is the Co-founder and CEO of Superluminal. He holds a Ph.D. in Applied Biology from the University of London and has held senior executive positions, including Chief Business Officer (CBO) and Chief Executive Officer (CEO), at Evotec OAI, Proteros, Evotec, IGEN, and Corning Incorporated. With over a decade of extensive experience in the biochemical and pharmaceutical industry, he has been involved in numerous venture capital financings and business development transactions within the pharmaceutical sector.

Ajay Yekkirala is another co-founder of Superluminal, serving as Senior Vice President and Head of Research and Development. He also serves as Vice President and Head of Biology at RA Capital Management, and is a co-founder and board member of Blue Therapeutics. A dynamic leader and biotechnologist, he brings over 17 years of industry and academic experience in life sciences drug development. He holds a Ph.D. from the University of Minnesota Medical School, completed a postdoctoral fellowship at Harvard Medical School, and has published 16 papers in the field of biomedicine as first or co-author.

Superluminal’s emergence as a rising star in the small-molecule drug discovery market is closely tied to its two co-founders. Cony D’Cruz currently serves as a Venture Partner at RA Capital Management, where she oversees the development of the firm’s drug screening platform. By integrating physics-based methods with machine learning (ML), the platform ultimately accelerates the development of candidate drugs.

During this process, Cony D’Cruz conceived the idea of practically implementing the platform technology by establishing an AI-centric drug discovery company. Thus, Superluminal was born. “Superluminal,” literally meaning “faster than light,” reflects the vision of Ajay Yekkirala and his team to accelerate new drug discovery, aiming to shorten the development timeline for clinical candidates from years to months.

Behind the acceleration of drug R&D is a strong scientific research team. Superluminal’s research team includes Yamina A. Berchiche, Ph.D. in Biochemistry from the University of Montreal and Founder and Director of Dr.GPCR; Yang-Ming Zhu, an AI engineer who co-holds 22 U.S. patents and has authored approximately 80 peer-reviewed papers; Marek Orzechowski, Ph.D. in Computer Science from the University of Warsaw; and Jiming Chen, Ph.D. in Chemical and Biomolecular Engineering from the University of Illinois. They possess extensive experience in artificial intelligence, computational chemistry, and small-molecule drug screening and development.

G protein-coupled receptors (GPCRs), also known as seven-transmembrane receptors (7TMRs), are composed of seven α-helical transmembrane domains and represent one of the most important families of integral membrane proteins. They constitute the largest protein family encoded by the human genome, comprising approximately 800–1,000 genes, and are widely distributed in organs and tissues such as the central nervous system, immune system, cardiovascular system, and retina.

Since GPCRs are all located on the cell membrane, have complex functions, and participate in the regulation of most physiological processes, including development, vision, olfaction, and taste, they are associated with disease areas such as cardiovascular diseases, AIDS, migraine, and cancer. Compared to other drug targets, GPCRs are one of the most extensively studied targets in drug discovery due to their diverse pathophysiological roles.

Since its initial discovery in the 1970s, it has become a major research focus for biotechnology companies and scientists. According to a report published by Alexander S. Hauser et al., as of 2017, there were 475 drugs on the market targeting 108 GPCR targets, accounting for approximately 34% of all FDA-approved drugs.

Although GPCRs are associated with various diseases and represent one of the most important targets in drug development, the field of GPCR-targeted drug discovery is characterized by high risk and high difficulty.

On the one hand, not all GPCR targets are suitable for drug development. Among the approximately 800 human gene-encoded GPCRs, there are about 400 olfactory receptors and 150 orphan receptors. On the other hand, due to limited drug selectivity and a lack of precise regulation over receptor signaling pathways, GPCRs are challenging targets. Furthermore, high clinical trial costs and relatively low success rates (below 6.2%) have constrained the development of GPCR-targeted drugs, with currently only 10% of GPCRs having targeted drugs under investigation.

Nevertheless, Superluminal continues to prioritize GPCR targets in its small-molecule drug development pipeline. This confidence stems from its proprietary drug discovery platform, which leverages AI as its core technology by integrating generative biology, chemistry, machine learning, and a specialized big data infrastructure. Through this platform, the Superluminal team can rapidly screen millions of compounds to identify potential clinical candidates for disease treatment.

The team led by Ajay Yekkirala leveraged this platform to create a “predict-design-test” architecture capable of accurately simulating protein shapes and determining protein conformations. Within four months, the team transitioned from structure prediction to in silico screening of a 10-billion-molecule library, generated candidate compounds with differentiated Target Product Profiles (TPPs), evaluated drugability using ADMET filters, and ultimately delivered a biologically validated proof-of-concept (POC) GPCR-biased agonist.

AI Technology Becomes the Core Method for Superluminal’s Small-Molecule Drug Development. Ajay Yekkirala stated that leveraging AI technology can shorten the development timeline for clinical drug candidates from several years to just a few months. According to Cony D’Cruz, Superluminal’s 10-member team completed the entire process—from selecting a protein target to designing a compound molecule targeting that protein—within five months, achieving the expected results in laboratory studies.

Currently, as the precise structures of most GPCR protein targets remain unresolved, it is difficult to conduct structure-based drug design for accurate modulation of GPCR function.

Superluminal Medicines has effectively addressed the above issues with its technology.

First, it is essential to highlight a core technology within AI: Free Energy Perturbation (FEP). FEP is a physics-based method used to determine the binding affinity between compound molecules and their protein targets by predicting the relative changes in binding free energy among congeners. In FEP calculations, molecules are gradually transformed into structurally closely related analogs via non-physical pathways. By calculating and evaluating the free energy difference (ΔΔG) between the end states of these transformations, prediction accuracy comparable to experimental values, with an error margin of approximately 1 kcal/mol, can be achieved.

In 2021, DeepMind released AlphaFold 2, a machine learning-based algorithm for protein structure prediction. AlphaFold 2’s predictions for most protein structures deviate from the actual structures by only the width of a single atom, achieving accuracy comparable to that of observations made using complex instruments such as cryo-electron microscopy.

According to the paper published on the Superluminal Medicines website, AlphaFold2-generated structures are sufficiently accurate under appropriate conditions to be utilized in the lead optimization phase of drug discovery projects via physics-based methods such as Free Energy Perturbation (FEP). However, AlphaFold2 models are not ideal for virtual screening; researchers should incorporate post-processing modeling to refine binding sites into more realistic holo-state models.

Based on this, Superluminal collaborated with Isaac Filella-Merce and colleagues from the Barcelona Supercomputing Center to develop a generative AI (GM) workflow. This GM can learn from molecular metrics, including drug-likeness, synthesizability, similarity, and docking scores.

The research team tested the generative model (GM) in two model systems, CDK2 and KRAS. In both cases, the model generated chemically feasible molecules with high predicted affinity for the targets, at proportions significantly higher than those observed in the training data. Furthermore, novel scaffolds distinct from the known scaffolds for each target were identified, opening up new possibilities for drug discovery. Meanwhile, Superluminal has enhanced its virtual drug screening technology to predict key characteristics such as absorption, distribution, metabolism, and excretion (ADME) properties, thereby helping to improve the performance of GPCR-targeted drugs in human clinical trials.

This also means that Superluminal can dynamically probe proteins to explore the multiple conformations they can adopt, thereby accelerating the determination of GPCR structures and their conformational changes.

Superluminal Medicines is expected to advance GPCR drug candidates in the near term. However, it has not yet disclosed details of the selected GPCR targets and the targeted disease areas, nor has it revealed the timeline for its first batch of candidates to enter clinical trials.

In 2001, imatinib (Imatinib, Gleevec), developed by Novartis, received FDA approval for the treatment of diseases such as chronic myeloid leukemia and gastrointestinal stromal tumors, becoming the world’s first approved small-molecule targeted therapy. The research and development of innovative small-molecule drugs has entered the era of targeted therapy and is advancing toward an age of greater precision.

GPCR-targeted drugs have become a key niche in the field of innovative small-molecule therapeutics. According to a report by Mordor Intelligence, the market size for small-molecule drug discovery is projected to grow from USD 52.35 billion in 2023 to USD 79.67 billion in 2028, representing a compound annual growth rate (CAGR) of 8.76% during the forecast period (2023–2028). Meanwhile, the global GPCR market is expected to increase from USD 5.71 billion in 2023 to USD 8.02 billion in 2028, with a CAGR of 7.01% over the same forecast period (2023–2028).

Among these, because GPCR signaling affects different aspects of cancer biology, such as invasion, migration, and vascular remodeling, GPCRs thus have the ability to regulate cellular functions that mediate cell survival, tumor progression, and metastasis, which contributes to cancer proliferation, angiogenesis, invasion, and metastasis.

According to data released by the International Agency for Research on Cancer, there were approximately 19.3 million new cancer cases and 10 million cancer-related deaths worldwide in 2020. The growing demand for effective therapies has made GPCRs and their signaling pathways—central to the new era of precision medicine and cancer immunotherapy—a key focus of current and future research.

Currently, the highly competitive GPCR landscape features several well-established players, including Structure Therapeutics, Domain Therapeutics, and Escient Pharmaceuticals. How will Superluminal Medicines carve out a market niche with its upcoming GPCR-targeted therapeutics? We will continue to monitor developments.

References:

[1]. Anna M. Díaz-Rovira, Helena Martín, Thijs Beuming, Lucía Díaz, Victor Guallar, and Soumya S. Ray. Are Deep Learning Structural Models Sufficiently Accurate for Virtual Screening? Application of Docking Algorithms to AlphaFold2 Predicted Structures. ACS Publications Most Trusted. Most Cited. Most Read. https://pubs.acs.org/doi/pdf/10.1021/acs.jcim.2c01270

[2]. Thijs Beuming, Helena Martín, Anna M. Díaz-Rovira, Lucía Díaz, Victor Guallar, and Soumya S. Ray. Are Deep Learning Structural Models Sufficiently Accurate for Free-Energy Calculations? Application of FEP+ to AlphaFold2-Predicted Structures. ACS Publications Most Trusted. Most Cited. Most Read. https://pubs.acs.org/doi/pdf/10.1021/acs.jcim.2c00796

[3]. saac Filella-Merce, Alexis Molina, Marek Orzechowski, Lucía Díaz, Yang Ming Zhu, Julia Vilalta Mor, Laura Malo , Ajay S Yekkirala, Soumya Ray, Victor Guallar. Optimizing Drug Design by Merging Generative AI With Active Learning Frameworks. https://arxiv.org/ftp/arxiv/papers/2305/2305.06334.pdf