New AI model EVA accelerates drug discovery in inflammation, bridging preclinical research to real-world patient therapies.
Drug discovery is always about patience, trial and error. For every promising compound identified in a lab, many fail to become effective therapies for humans. In immunology, this challenge is even greater. Scienta Lab, a Paris-based deeptech startup, is launching a new multimodal AI model called EVA, designed to guide researchers through the maze of preclinical and clinical data, helping them focus on the drug candidates most likely to succeed [1].
Julien Duquesne, CTO and co-founder of Scienta Lab, explains that “the goal is not to replace experimentation, but to better guide decision-making at every stage of the development.” In other words, EVA isn’t a replacement for scientists, but a co-pilot helping them make smarter choices faster.
At its core, EVA is like a multilingual translator – but for biology. It reads data from multiple sources: gene activity, tissue samples and protein levels, across humans and lab animals. By harmonizing this information, EVA can predict how a drug candidate might perform in humans before trials even begin.
Think of it like mapping a city with incomplete satellite images. Traditional methods give you fragments of streets and buildings; EVA connects those fragments into a complete, understandable map. Researchers can then decide which paths are worth exploring and which are dead ends.
This is especially relevant for “inflammaging,” the low-level, chronic inflammation that quietly drives many age-related diseases like arthritis, diabetes and Alzheimer’s. By identifying the molecular pathways that cause inflammation, EVA could help accelerate therapies that target not just symptoms, but the underlying mechanisms of aging itself.
Benchmarks suggest EVA performs up to twice as well as current AI models in predicting which drug targets will succeed. This is more than just a technical feat since it translates directly into fewer late-stage failures, faster clinical trials and ultimately quicker access to treatments for patients.
For a field like longevity science, where every year matters, this could shift timelines from decades to years. Imagine getting actionable insights on how to reduce chronic inflammation years earlier than current methods allow. This is the kind of progress EVA promises.
Scienta Lab is taking a dual approach: it has released an open version of EVA’s transcriptomic model, allowing researchers worldwide to explore immune-mediated diseases with a powerful new tool. At the same time, large-scale, customized applications are offered through commercial partnerships.
Openness and practical deployment are a model for the longevity sector. Collaboration can drive innovation, while commercial focus ensures therapies actually reach the patients who need them.
Longevity.Technology: EVA represents a step toward precision longevity medicine. Chronic inflammation is one of the key drivers of age-related decline, and targeting it effectively could extend healthy lifespan. Tools like EVA give researchers a head start, helping them navigate the complexity of biology with insight and confidence rather than guesswork.
In a broader sense, EVA is emblematic of the progressive shift in longevity research: integrating cutting-edge technology, open science and translational biology to tackle aging not as an inevitability, but as a challenge we can actively manage. As the field continues to evolve, AI models like EVA could become essential allies in the quest to slow aging, prevent disease and expand healthspan.
