Researchers at the University of California San Diego have developed artificial intelligence models and digital cell twins to track how shifting mitochondrial networks respond to drugs, according to studies published in the journal Cell.
Tracking Mitochondrial Networks With 4D Lattice Light-Sheet Microscopy
Textbooks typically depict mitochondria as static, bean-shaped structures inside cells. In reality, they operate as a shifting power grid that constantly merges, splits, and moves to areas requiring energy, according to the University of California San Diego research team. This continuous movement has historically made large-scale study difficult.
To capture this activity, researchers utilized 4D lattice light-sheet microscopy. This technology tracks mitochondria and other cellular structures in 3D space over time. Using these measurements, the team gathered 40,000 single-cell 4D movies of cancer cells treated with 25 compounds that disrupt mitochondria in different ways.
MitoSpace AI and Rules-Based Digital Twins
The research project centered on two distinct virtual systems designed to measure how mitochondrial networks change. According to the University of California San Diego team, the first tool is MitoSpace, a deep-learning system that interprets mitochondrial form as a readout of cell health.
Instead of relying on manual sorting, MitoSpace automatically detected patterns and grouped cells by their drug responses without prior labeling of individual treatments. The second tool is a physics-based, rules-based digital twin that functions as a simulation of a living cell. When tested with drugs, the simulated mitochondrial networks reacted similarly to those observed in real cells, pointing toward potential in silico testing before laboratory bench work.
Implications for Drug Screening and Disease Research
Mitochondrial shape changes signal whether a cell is healthy, stressed, or damaged. Because of this, researchers frequently use them as markers when studying cancer, diabetes, Alzheimer’s, and mitochondrial disorders in children. Traditional lab imaging often captures only a single flat frame, missing the continual remodeling that informs treatment response.
According to the published findings, these digital tools could accelerate drug screening by allowing scientists to test ideas virtually before committing to slower, more expensive laboratory work. The current methodology remains an early-stage research tool rather than a consumer technology, aimed at identifying promising compounds and evaluating their cellular impacts more efficiently than static imaging permits.
Did You Know?
Mitochondria do much more than generate energy. Their physical shape constantly shifts to signal cellular stress, making them vital markers for researchers investigating complex conditions like Alzheimer’s and cancer, according to data published in Cell.

Frequently Asked Questions
What are mitochondrial digital twins?
According to research from the University of California San Diego, digital twins are physics-based, rules-based simulation models that mimic how organelles behave inside a real living cell.
How does MitoSpace work?
MitoSpace is a deep-learning system that analyzes mitochondrial shapes captured via 4D lattice light-sheet microscopy to evaluate cell health and group cells by their drug responses automatically.
Is this technology available for consumer use?
Want to stay updated on the latest breakthroughs in biotechnology and artificial intelligence? Subscribe to our free newsletter today for smart advice, easy tips, and exclusive updates.
Related reading