AI and Multi-Omics Join Forces to Rethink Drug Discovery

Athos and Xeptiva use AI-powered multi-omics to sharpen drug discovery and reduce costly development risks

13 Feb 2026

Two biotechnology companies are seeking to reduce the high failure rate of experimental medicines by embedding artificial intelligence at the core of drug discovery.

Athos Therapeutics and Xeptiva Therapeutics have formed a partnership to integrate AI-driven analysis into early research. Their aim is to identify stronger drug targets sooner and limit the costly setbacks that often occur in later clinical stages.

Central to the collaboration is AthosOmics.AI, a cloud-based platform designed to combine genetic, proteomic, metabolic and clinical data within a single system. Rather than analysing each layer of biology separately, the platform maps them together to detect patterns that may otherwise be overlooked.

The companies say the approach can help clarify disease pathways, identify biomarkers and prioritise high-confidence targets before programmes advance into expensive trials. About 90 per cent of experimental therapies fail to secure regulatory approval, often after years of development and significant capital outlay.

Athos has already applied the system to its investigational therapy, ATH-063, using integrated data analysis to better understand the treatment’s underlying mechanisms and refine its development strategy. Within the partnership, Xeptiva is progressing a veterinary therapeutic vaccine, extending the use of AI-enabled analytics beyond human medicine into animal health.

The collaboration reflects a wider shift across the pharmaceutical industry. AI tools, once viewed as experimental add-ons, are increasingly being incorporated into core research workflows. Companies face rising development costs and growing investor pressure to demonstrate clearer, data-backed paths to value.

Platforms capable of synthesising large biological datasets promise more disciplined decision-making and tighter allocation of capital. For smaller biotechnology groups, scalable AI infrastructure may also narrow the gap with larger pharmaceutical companies that have traditionally dominated research budgets.

Regulators are still refining standards for validating AI-enabled tools in drug development, and questions around transparency and reproducibility remain. As multi-omics datasets expand, integrated analytics are likely to become a routine expectation rather than a competitive advantage.

Whether such systems can materially improve approval rates will become clearer as AI-informed programmes move through clinical testing.