How BluMaiden uses AI to transform small-molecule drug discovery

BluMaiden Biosciences, a Singapore-based biotech company working in drug discovery, recently announced that it has secured “substantial investment” in a round led by deep tech VC firm Elev8.vc and joined by SEEDS Capital.

The company has developed what they refer to as a transformative approach to small molecule drug discovery, addressing a
significant challenge in the pharmaceutical industry: the limited diversity of chemical compounds in traditional chemical libraries, which restricts the scope of potential drug candidates and hinders innovation in drug discovery.

BluMaiden’s solution involves probing the vast chemical space hidden within the human body, using advanced AI-guided computational genetics and chemistry.

The company’s co-founders are Rohan Williams, Michael Tillmann, and Damien Keogh.

Rohan Williams, Ph.D., is a scientific co-founder at BluMaiden with expertise in bioinformatics and systems biology. He is also the Head of the SCELSE Integrative Analysis Unit. Michael Tillmann, a founding member and Chairman of the Board, has been instrumental in BluMaiden’s growth, leveraging his experience as a former CEO at Roche Diagnostics.

Also Read: Asia’s biotech boom: Innovation, investment, and a new era of discovery

Before founding BluMaiden, Keogh was actively engaged in the biotech sector, including roles in venture capital as an entrepreneur-in-residences, involved in spin-out and venture-created biotech companies, and being an alumnus of Johnson & Johnson Innovation JLABS.

In this email interview with e27, Keogh explains the problems the startup aims to solve and how AI technology helps it achieve these goals.

This is an edited excerpt of the conversation.

What specific problem or need does BluMaiden address, and what inspired you to take on this challenge?

The pharmaceutical industry encounters significant challenges with chemical libraries in small-molecule drug discovery. One key issue is repeated rediscovery of known compounds, which wastes resources without yielding new treatments.

Additionally, these libraries often have molecular biases that limit compound diversity, reducing the likelihood of finding innovative drugs. Since the early 1980s, about 50 per cent of successful FDA-approved small-molecule drugs have been derived from natural products, owing to their structural diversity and biological activities. Nearly half of these are synthetic drugs that mimic natural products.

My motivation to establish BluMaiden comes from my research interest in metabolite biology, which focuses on the mechanisms and effects of small molecules derived from natural products. At BluMaiden, our diverse team explores new chemical spaces within the human body, linking this with clinical data to uncover pharmacological significance.

How does your AI-powered platform integrate computational genetics and chemistry to enhance the drug discovery process? Can you share a specific example of a breakthrough or significant finding your technology has facilitated in drug discovery?

Our AI-guided technology integrates computational genetics and chemistry to improve drug discovery, and we begin by harnessing natural products as a rich reservoir for new small-molecule drug candidates. Our team brings expertise across various scientific and technological domains.

Also Read: Biotech co-working provider NSG BioLabs concludes US$14.5M financing round 

And we have become quite effective at communicating and understanding across these disciplines. We utilise customised machine learning models and advanced feature engineering to extract health-predictive signals from microbiome and clinical data.

We identify novel therapeutic opportunities by analysing long-range semantic relationships within extensive and diverse datasets.

What are the biggest challenges you face in developing and deploying AI algorithms for drug discovery, and how do you tackle them?

We believe that starting from clinical evidence gives us a unique edge and a higher chance of finding safe and effective drugs. This is why we’ve assembled a team of diverse science and technology experts to explore novel chemistry space within the human body, a natural product, and link this with clinical data (human evidence) to provide pharmacological significance.

Obtaining reliable clinical data in sufficient quantity is a crucial challenge for many biotechs, including our early-stage drug discovery programmes. BluMaiden addresses this through our Pharma Services division.

By offering comprehensive services, we aim to gain early access to clinical studies and ensure reliable data generation through a regulatory-compliant, pharma-grade quality management system. This data feeds directly into our drug discovery engine.

The synergistic relationship between these divisions ensures smooth data flow and sustainability for our business.

What is your company’s business model? Who are your users, and how do you acquire them?

We are prioritising sustainability and profitability; from the beginning, we have committed to building a sustainable business model. Despite making some strategic adjustments, we have, for example, successfully developed a globally deployed Pharma Services for clinical trial analytics. As an auditable vendor to clinical trial sponsors, we deliver comprehensive analysis and reporting, including patient stratification, drug responder vs non-responder analysis, end-point optimisation, and more.

Also Read: Forte Biotech: Helping farmers with early detection of prawn diseases in Vietnam

With Amazon AWS, we are recognised as an AWS Qualified Software Provider and have deployed this in key markets in the US and Europe.

What role do partnerships and collaborations play in your approach to leveraging AI for drug discovery?

Partnerships and collaborations are a cornerstone of BluMaiden’s strategy. We’ve established several global partnerships with tech and life science companies to leverage our capabilities to key markets in the USA and Europe. These partnerships are particularly significant for our Pharma Services, enhancing clinical trial decision-making.

These partnerships are also a channel for obtaining reliable and ample clinical data crucial for our early-stage drug discovery programs. By offering comprehensive services, we gain early access to clinical data and ensure high-quality, regulatory-compliant data generation. This data is integrated into our drug discovery engine, fostering a synergistic relationship between our drug discovery division and Pharma Services, promoting efficient data flow and business sustainability.

What are the most exciting developments or future applications you anticipate for your AI-powered drug discovery platform?

There are many exciting prospects, but what excites me the most is the ability to predict previously undiscovered natural products and their signalling targets with a high success rate. This advancement will significantly expand chemical diversity and increase the odds of finding drug-like candidates.

What is your big plan for 2024 and beyond?

Over the past few years, we have built a solid foundation in our R&D and commercial capabilities. In 2024, we aim to significantly advance our drug discovery programs and revenue pipelines through our engagements with Pharma. We aim to establish a self-sustaining venture model for our biotechnology company and demonstrate to our shareholders the benefits of risk-adjusted returns on their investments.

—

Image Credit: BluMaiden

This article was first published on July 30, 2024

The post How BluMaiden uses AI to transform small-molecule drug discovery appeared first on e27.