Dear Readers,

Welcome to the inaugural issue of BioUpdates.AI! We are thrilled to have you join our community of professionals who are passionate about the transformative potential of AI in life sciences. Whether your interest is in drug discovery and development, business processes and internal functions, or engagement and marketing, we have you covered in terms of bringing you the latest news and analysis where life sciences and AI meet.

Our mission with BioUpdates.AI is to help you stay informed about the latest trends, breakthroughs, and innovations that are transforming the life sciences industry as well as the policy, ethics, and regulatory developments that influence how AI tools are deployed across our sector. Each issue will deliver curated content designed to keep you ahead of the curve.

For now, you can expect to receive our newsletter twice a month. We look forward to your feedback and engagement.

Thank you for being a part of BioUpdates.AI!

Warm regards,

The BioUpdates.AI Team and Sigla Sciences

How artificial intelligence is transforming drug discovery and development

One of the hottest questions in biotech today is whether AI will be fundamentally transformative in drug discovery. The thesis is that AI will allow us to better leverage a combination of vast amounts of data (including almost everything that falls within the “-omics” data category), machine learning algorithms, and computational power to quickly uncover new discoveries and predict what molecules will work for what scenarios. While this may be possible, many in the industry are also wary of technology fads, particularly those that carry a high initial investment and operational overhead in terms of costs (servers aren’t cheap!). On the other hand, a few unicorn-level companies have already made their mark on this space and are actively looking for partnering/BD&L opportunities with molecules in their pipelines. What are the opportunities, and what are the tradeoffs?

Traditional drug discovery methods are often time-consuming and costly, with a high rate of failure. AI promises to change this paradigm by accelerating identification and optimization of drug candidates; it can also potentially help on the other side of the equation as well, in terms of finding new or better targets. AI can also help us better predict how candidates and targets will interact, and where we may get undesirable off-target effects. All of these are useful features.

Ok, so what’s the catch?

Where we have high potential, we also have challenges. In the midst of the hype, it’s often unpopular to talk about the forces that might oppose this seeming revolution:

• Data Quality and Bias: Any AI engine or tool is completely reliant on data. Poor data quality creates a quintessential Garbage-in-Garbage-out problem. Humans may do a poor job of selecting sufficiently unbiased datasets as AI inputs (or may not know how to evaluate datasets for bias). As an even thornier challenge, humans have a very difficult time assessing the bias of data and algorithms after the fact; AI can be a bit of a black box. Bad data can lead to dangerous outcomes and untrusted AI outputs. In drug discovery, this could lead to expensive program mistakes and misses.

• Cost: Implementing AI tech can be expensive, requiring investment in infrastructure, talent, and maintenance. Smaller biotech firms or those in developing regions may find it challenging to adopt these technologies, potentially widening the gap between larger and smaller entities in the industry. On the other hand, those who buy wisely could catapult ahead of even legacy players because of potential efficiencies; this is a tradeoff not to be taken lightly. On top of these considerations, executives and managers will be flooded with sales meeting requests from AI companies riding the hype and offering new solutions, many of which may not deliver on their promises. Executives need to be discerning when evaluating these offerings.

• Regulatory Hurdles: The regulatory landscape for AI in biotech is still evolving. Navigating the complex regulatory requirements and ensuring compliance will be daunting. Much remains unknown — creating more risks for those who are pushing forward.

Who’s doing it?

Here’s a look at a handful of the companies shaping the landscape in AI and Drug Discovery, and what they are doing.

• Deep Genomics: Uses an in-house tool “AI Workbench” in RNA-based therapy discovery/development for rare genetic diseases.

• Recursion Pharmaceuticals: Their platform identifies novel treatments by analyzing biological image data at scale.

• Insitro: With an all-start veteran team of biotech leaders on its Board and a cool $400M+ raised, this “quiet unicorn” is using machine learning to shorten development timelines and lower discovery costs.

• Atomwise: Uses AI to predict the binding affinity of small molecules to targets, aiming to speed up identification of drug candidates.

• BenevolentAI: Uses its “BenAI” toolset to derive insights for drug discovery and molecule repurposing from hundreds of data sources including scientific and regulatory documents and “-omics” datasets.

• Exscientia: Developing candidates in oncology, hematology, and inflammatory disease based on AI-derived targets and molecules.

• Cyclica: Recently acquired by Recursion, Cyclica focused on the proteome, “low data targets,” and pharmacology of small molecules.

• Insilico Medicine: Developed the Pharma AI software suite for target discovery, novel molecule generation, and design/predict for clinical trials.

• BPG Bio: Developing a hematology/oncology-heavy pipeline derived from their “Interrogative Biology” AI-based platform and a large biobank from academic research and hospital partners.

• Healx: UK-based AI drug discovery company with a special focus on rare diseases and predicting new uses or combinations of existing drugs.

• Aitia: Leveraging digital twin technology to build computational models of disease from multi-omic patient data.

• BioSymetrics: Using predictive models to downrisk clinical development with a focus on neurological, cardiometabolic and rare diseases.

What can we expect?

The opportunities AI presents for drug discovery are significant, but navigating the landscape requires a balanced approach; executives need a clear-eyed view and thorough vetting of potential partnerships. AI has an undeniable allure. We all want faster, better drugs with more predictable clinical results. But we’re still in early days.

Ultimately, the integration of AI into drug discovery is not a question of if, but when and how effectively it can be done. Companies such as those listed above will be bellwethers for wider acceptance and partnership potential with Big Pharma dollars at stake.

Innovation Pulse for Aug. 5, 2024: AI & Life Sci in the News

Tracking the latest breakthroughs and policy shifts shaping the future of biotech and AI

• Paige publishes open-source AI pathology models for cancer R&D

• Healx, an AI-enabled drug discovery platform for rare diseases, raises $47M

• Rejuvenate and SAS partner to harness AI for age-related diseases

• World’s first major AI law enters into force

• A CIO canceled a Microsoft AI deal. The reason should worry the entire tech industry