Deep Analysis of Mitochondria and Cell Health Using Machine Learning
Atena Zahedi 1,3, Vincent On 2, Rattapol Phandthong 3, Angela Chaili 3, Guadalupe Remark 3, Bir Bhanu 1,2,4, Prue Talbot 1,3,✉
Mitochondria are often called the “powerhouse of the cell”, but in reality, they are much more than tiny batteries. These dynamic structures produce ATP – the energetic currency of the cell, respond to stress and can influence whether a cell survives or dies.
Since live mitochondria are constantly moving and changing shape, they can be difficult for scientists to study. This is where MitoMo, a mitochondrial analysis software, really shines. This software uses machine learning to analyze microscope images and videos of mitochondria. This allows scientists to study their shape, motion and behaviour over time. In this study, MitoMo could distinguish between different mitochondrial forms and dictate how they changed during stress.
This is really important because mitochondrial dysfunction is linked to diseases like cancer, metabolic disorders, and even neurodegenerative diseases. Tools like MitoMo can help researchers understand cellular health through a mitochondrial lens. Even better, these researchers have made their software ‘open source’, meaning that more labs have the freedom to use this tool. As open-source tools like MitoMo become more widely available, they could help researchers uncover new links between mitochondrial health and human disease.
