The National Institute of Neurological Disorders and Stroke, part of the National Institutes of Health (NIH), has awarded a $3.9 million grant to a multi-institutional team led by scientists at Sanford Burnham Prebys Medical Discovery Institute to develop a non-opioid pain therapeutic and advance it into a Phase 1 clinical trial.

A Virginia Tech research scientists with the Fralin Biomedical Research Institute at VTC, has been awarded a National Institutes of Health Mentored Research Scientist Development Award to study how metabolic health influences the brain systems that shape what and how we eat. The four-year award will support her research into how obesity and insulin resistance affect reward learning and decision-making in the human brain. 

* The project will use environmental DNA (eDNA)—traces of genetic material left behind by organisms—to understand the health of streams throughout California and assess impacts from land use and climate change.

* This work will combine advanced machine learning and geospatial data with the on-the-ground efforts of hundreds of volunteer community scientists.

* The project will launch an open-source, low-cost cloud platform so that Indigenous tribes, land managers, watershed groups, and local agencies have the tools to evaluate stream conditions and monitor biodiversity.

A research team led by Virginia Tech cybersecurity expert Bimal Viswanath has found a critical blind spot in today's image protection techniques designed to prevent bad actors from stealing online content for unauthorized artificial intelligence training, style mimicry, and deepfake manipulations.

Insilico Medicine announced that its research paper, “When Single Answer Is Not Enough: Rethinking Single-Step Retrosynthesis Benchmarks for LLMs,” has been accepted for presentation at the International Conference on Machine Learning 2026. The study challenges conventional retrosynthesis benchmarking approaches that rely on single “ground-truth” answers and Top-K accuracy metrics, which may not reflect the multi-solution nature of real-world chemistry.

The paper introduces ChemCensor, a chemistry-aware evaluation metric designed to assess model performance based on reaction centers and functional groups, aligning more closely with expert human reasoning. Additional contributions include the CREED dataset, comprising 6.4 million validated reactions; benchmarking results from the C3LM model; and the URSA-expert-2026 dataset, an expert-annotated benchmark designed to reduce data leakage and improve evaluation rigor.

The research supports the development of more realistic and scalable training and evaluation frameworks for AI-driven retrosynthesis and drug discovery. Supporting materials will be made publicly available to promote transparency and reproducibility.