Key Highlights
- Shift Bioscience proposes a new framework to improve gene target discovery for anti-aging drugs
- Study shows current virtual cell models often perform no better than dataset averages
- New methods aim to boost biological relevance and accelerate rejuvenation R&D
Deeper Dive
Cambridge-based Shift Bioscience is rethinking how AI models are evaluated for rejuvenation research. Their study, led by Lucas Paulo de Lima Camillo, reveals that many virtual cell models — which simulate gene expression changes for drug discovery — fall short by merely echoing the dataset average, limiting their biological usefulness.
To fix this, the team introduced new strategies like DEG-weighted scoring and baseline calibrations, helping prioritize models that better predict real, meaningful gene perturbations. This shift could dramatically accelerate discovery pipelines in rejuvenation therapeutics by reliably identifying new targets for age-related interventions.
“By improving these metrics and baselines, we can build more powerful perturbation models and speed up the search for anti-aging drug targets,” de Lima Camillo explained.
Shift Bioscience continues to advance AI-driven approaches to reverse the biology of aging, strengthening its mission to fight age-related disease through smarter gene target discovery.
About Shift Bioscience
Shift Bioscience is a Cambridge-based biotechnology company pioneering the use of AI-driven gene discovery to develop rejuvenation therapeutics targeting the root causes of aging and age-related diseases. By combining advanced machine learning with single-cell technologies, Shift Bioscience aims to unlock new gene targets and accelerate the path toward effective anti-aging therapies.
By