Predicting Stem Cell Bone Growth: Metabolomics and AI Team Up

Scientists often use stem cells to make artificial bones, but making sure these cells develop properly can be tricky. They've tried using omics technologies, like metabolomics, to check their quality. In a recent study, researchers combined metabolomics with machine learning to see if it could tell the difference between stem cells that became bone-like and those that didn't. They found 11 metabolites, chemical fingerprints, that helped train a model to spot the difference. The model was really good at picking out bone-like cells in 3D cultures, but not so great with cells from Wharton's Jelly, which are known for not becoming bone very well. They found that certain metabolites, like fumarate, glycerol, and myo-inositol, were key signs of bone growth. These findings could help create better tests to ensure artificial bone grafts work as they should.