Finding New Drugs Is Laborious. Potential Sciences, a startup backed by the UK government, is betting that training AI models on computer-generated biochemical data can speed up the plodding process of developing new drugs. Drug discovery is a needle in a haystack problem. This AI startup wants to simulate the haystack to speed the process up.Photo-Illustration: Jobanny Cabrera; Getty Images A new British startup believes that artificial intelligence is the key to speedrunning the slow and temperamental process of drug design. Potential Sciences plans to build large-scale chemical datasets by using basic principles of physics to simulate interactions between atoms, then small molecules, then complex amino acids and proteins.

In breathless press releases and posts, the company has touted its models' ability to deliver scientific breakthroughs. But its IPO filing taps the brakes on just how fast its AI will lead to big discoveries. AI giant Anthropic and CEO Dario Amodei have repeatedly claimed that the company's models will revolutionize biology and drug discovery . But in its IPO filing Monday, Anthropic struck a humbler note, acknowledging its lack of experience in physical experimentation and the potential pitfalls of drug development. Wherever it's going, it seems, it will take some time to get there.

As generative AI captures public attention, a different kind of AI is reshaping drug discovery. Machine learning models are helping to compress decade-long timelines and cracking problems that were previously unsolvable. Designing and developing a new medicine is an expensive, failure-prone scientific challenge. A new drug can take many years to develop, at the cost of a significant investment. And even then, most possible candidates never reach the patient. For biologic medicines, therapies made from engineered proteins rather than synthetic chemistry (which are often used to treat conditions across most major acute and chronic diseases), the complexity is even greater.

Check your subscription status, update your details and more. Drug development must change too. In this New Scientist CoLab podcast, experts from global life sciences leader Cytiva explain the hidden, high-stakes science of purification that is required to close the gap between drug discovery and the pharmacy shelf. Artificial intelligence and big data are flooding discovery pipelines with high-potential drug candidates, but this rapid innovation has created a new challenge. Simply put, our capability to design miracle molecules is vastly outstripping our technology to mass-manufacture them safely for the global public.

The Download: Anthropic launches Claude Science, and California's carbon manure math Plus: The US has lifted restrictions on Anthropic's Mythos and Fable models. Claude Science is Anthropic's newest flagship product At an event for pharmaceutical executives, biotech founders, and researchers yesterday, Anthropic announced Claude Science, a major new product intended to support scientific research like Claude Code supports software engineering. Like Claude Code, Claude Science can autonomously carry out meaningful work from concise, high-level instructions, with tools for computational biology and drug development. The launch signals that Anthropic is doubling down on AI for science, and the company will also use the product in its own research into drugs for rare, neglected diseases. Discover why Anthropic is betting big on AI for scientific research . Why California's carbon manure math doesn't add up Years ago, the state set up a system that pays cattle farmers to turn the methane emitted from cattle manure into natural gas.

Predicting the bioactivity of candidate ligands remains a central challenge in drug discovery. Ligands and endogenous substrates often compete for the same binding sites on target proteins, and the extent to which a ligand can modulate protein function depends not only on its binding but also on how effectively it occupies the relevant pocket. However, most existing methods focus narrowly on local interactions within protein-ligand complexes and neglect spatial emptiness--the unoccupied regions within the binding site that may permit endogenous molecules to engage or interfere. Such unfilled space can diminish the ligand's functional impact, regardless of binding affinity. To overcome this key limitation in protein-ligand modeling, we propose LigoSpace, a novel method integrating three core components. LigoSpace introduces GeoREC (Geometric Representation of Spatial Emptiness in Complexes) to quantify atomic-level empty space and UnionPocket to unify multiple protein pockets, providing a global view of binding sites. Additionally, LigoSpace employs a pairwise loss instead of commonly used MSE loss, to better capture relative relationships critical for drug discovery. Extensive experiments on multiple datasets with diverse bioactivity types demonstrate that LigoSpace significantly improves performance when integrated into state-of-the-art models, highlighting the effectiveness of its novel components. Equal contribution, may cite either first.

Virtual Screening (VS) is vital for drug discovery but struggles with low hit rates and high computational costs. While Active Learning (AL) has shown promise in improving the efficiency of VS, traditional methods rely on inflexible and handcrafted heuristics, limiting adaptability in complex chemical spaces, particularly in balancing molecular diversity and selection accuracy. To overcome these challenges, we propose GLARE1, a reinforced active learning framework that reformulates VS as a Markov Decision Process (MDP). Using Group Relative Policy Optimization (GRPO), GLARE dynamically balances chemical diversity, biological relevance, and computational constraints, eliminating the need for inflexible heuristics. Experiments show GLARE outperforms state-of-the-art AL methods, with a 64.8% average improvement in Enrichment Factors (EF). Additionally, GLARE enhances the performance of VS foundation models like DrugCLIP, achieving up to an 8-fold improvement in EF0.5% with as few as 15 active molecules.

Predicting interactions between biochemical entities is a core challenge in drug discovery and systems biology, often hindered by limited data and poor generalization to unseen entities. Traditional discriminative models frequently underperform in such settings. We propose BioCG (Biochemical Constrained Generation), a novel framework that reformulates interaction prediction as a constrained sequence generation task. BioCG encodes target entities as unique discrete sequences via Iterative Residual Vector Quantization (I-RVQ) and trains a generative model to produce the sequence of an interacting partner given a query entity. A trie-guided constrained decoding mechanism, built from a catalog of valid target sequences, concentrates the model's learning on the critical distinctions between valid biochemical options, ensuring all outputs correspond to an entity within the pre-defined target catalog. An information-weighted training objective further focuses learning on the most critical decision points. BioCG achieves state-of-the-art (SOTA) performance across diverse tasks, Drug-Target Interaction (DTI), Drug-Drug Interaction (DDI), and Enzyme-Reaction Prediction, especially in data-scarce and cold-start conditions.

Designing protein binders targeting specific sites, which requires to generate realistic and functional interaction patterns, is a fundamental challenge in drug discovery. Current structure-based generative models are limited in generating nterfaces with sufficient rationality and interpretability.

Virtual Screening (VS) is vital for drug discovery but struggles with low hit rates and high computational costs. While Active Learning (AL) has shown promise in improving the efficiency of VS, traditional methods rely on inflexible and handcrafted heuristics, limiting adaptability in complex chemical spaces, particularly in balancing molecular diversity and selection accuracy. To overcome these challenges, we propose GLARE, a reinforced active learning framework that reformulates VS as a Markov Decision Process (MDP). Using Group Relative Policy Optimization (GRPO), GLARE dynamically balances chemical diversity, biological relevance, and computational constraints, eliminating the need for inflexible heuristics. Experiments show GLARE outperforms state-of-the-art AL methods, with a 64.8% average improvement in Enrichment Factors (EF). Additionally, GLARE enhances the performance of VS foundation models like DrugCLIP, achieving up to an 8-fold improvement in EF$_{0.5\\%}$