Curie Brief
Turn on cookies to sign in
Signing in saves your progress to your Curie account. We can only do that with cookies on — turn them on to continue.
News
Scientists at UC San Diego have built "virtual cells" — digital models that mimic real cell biology — to speed up drug discovery. Using 4D microscopy and AI, they can predict how cells respond to drugs based on mitochondrial shape and movement alone. The technology could fast-track treatments for cancer, diabetes, Alzheimer's, and rare mitochondrial disorders.
Researchers at UC San Diego have developed two types of "virtual cells" — sophisticated digital models that replicate the dynamic behavior of real cells — with the potential to dramatically accelerate drug discovery. Both approaches rely on 4D lattice light-sheet microscopy, which captures how cellular structures like mitochondria move in three dimensions over time, offering a far richer picture than the flat, static images traditionally used in drug screening.
The first tool, an AI model called MitoSpace, was trained on 40,000 4D movies of drug-treated cancer cells. Without any manual labeling, it learned to group cells by how they respond to drugs and predict a cell's energy state from mitochondrial shape alone. The second approach created a physics-based "digital twin" of a living cancer cell, accurately simulating how mitochondria behave — and even predicting their response to a drug it had never been tested on.
By the Numbers:
Why it matters: These virtual cell tools could reduce the need for time-intensive lab experiments, cut costs in early-stage drug development, and open new avenues for treating diseases like cancer, Alzheimer's, diabetes, and pediatric mitochondrial disorders — potentially reshaping how medicine is discovered.
