AGENTIC-SYSTEMSMay 16, 20261 min read7 views

Finding the molecular switches behind new infectious diseases

Professor Clare Bryant at the University of Cambridge is using the AI tool Co-Scientist to identify molecular switches behind zoonotic diseases, accelerating the research process from years to months.

The majority of emerging infectious diseases are caused by pathogens that jump from animals to humans, such as Ebola, HIV, flu and Covid-19. Professor Clare Bryant at the University of Cambridge is using Co-Scientist to hunt for the molecular switches that cause severe diseases, like sepsis, in humans when pathogens leap between species, and find new approaches to prevent this happening.

Testing Co-Scientist, Bryant fed it a summary of one of her grant proposals studying flu in birds and humans outlining her lab’s research questions. The tool generated and ranked a set of promising hypotheses – some she'd already considered, some she hadn't. The unfamiliar ones were the most thought-provoking.

When the grant was funded, Bryant fed in the full, detailed proposal. Later, reading through the output on a train to Brussels, she had an "a-ha!" moment: Co-Scientist had prioritised a protein that hadn’t been on her radar, connected to several signalling pathways she was already interested in. She spent the rest of the week itching to give it more data.

Back at her lab, she added unpublished material, kept confidential within Co-Scientist. With each back-and-forth, the hypotheses sharpened, moving from candidate proteins down to the specific amino acids her lab could focus its experiments on.

Bryant's team is now building cell lines containing the amino acid mutations to test the refined hypotheses. To get to the point of identifying precise amino acids would normally have taken two to three years of experimental work. But her lab is now on track to complete it in six months, she says, if the work with Co-Scientist has led them to the right targets.

Source: Google DeepMind Blog

More in this category

agentic-systems

Generative Ontology Induction: Domain-Agnostic Schema Discovery from Document Corpora Using Large Language Models

Researchers introduce Generative Ontology Induction (GOI), a domain-agnostic framework that automatically extracts structured ontologies from document corpora using LLMs. Achieving 95-100% structural coverage, GOI addresses a major bottleneck in knowledge-intensive AI systems.

agentic-systems

JUMP: Single-Pass Membership Inference on Fine-Tuned Diffusion Language Models

Researchers have proposed JUMP, a novel single-pass membership inference attack designed for fine-tuned discrete diffusion language models (dLLMs). By leveraging the unique properties of dLLMs, JUMP significantly improves detection accuracy while drastically reducing the number of required queries.

agentic-systems

Masked Diffusion Language Models are Strong and Steerable Text-Based World Models for Agentic RL

Researchers demonstrate that Masked Diffusion Language Models (MDLMs) serve as highly effective, steerable text-based world models for agentic reinforcement learning. By leveraging bidirectional denoising, MDLMs outperform autoregressive models four times their size in coherence, groundedness, and rollout diversity.

NOW LET US Related – Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment

agentic-systems

Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment

A new study demonstrates that small language models (SLMs) under 3 billion parameters can serve as highly capable local experts for specialized tasks. By combining structured benchmarking with low-cost parameter-efficient fine-tuning (PEFT), institutions can achieve AI autonomy without relying on expensive hardware.

agentic-systems

PlanFlip: Attacking Multi-Agent LLM Systems via Planning-Phase Prompt Injection

Researchers have introduced PlanFlip, a novel prompt injection attack framework targeting the planning phase of multi-agent LLM systems. The study reveals critical security blind spots in homogeneous agent pipelines and demonstrates that reasoning-augmented models like DeepSeek-R1 exhibit strong resistance.

agentic-systems

Rater State Bias in RLHF Preference Data: An Audit Framework

A new study identifies a structured bias in Reinforcement Learning from Human Feedback (RLHF) caused by the psychological state of human raters. Under stress, raters' shifting preferences can propagate through reward modeling, potentially compromising AI policy optimization.