Machine Learning of Hematopoietic Stem Cell Divisions from Paired Daughter Cell Expression Profiles Reveals Effects of Aging on Self-Renewal

Changes in stem cell activity may underpin aging. However, these changes are not completely understood. Here, we combined single-cell profiling with machine learning and in vivo functional studies to explore how hematopoietic stem cell (HSC) divisions patterns evolve with age. We first trained an artificial neural network (ANN) to accurately identify cell types in the hematopoietic hierarchy and predict their age from single-cell gene-expression patterns. We then used this ANN to compare identities of daughter cells immediately after HSC divisions and found that the self-renewal ability of individual HSCs declines with age. Furthermore, while HSC cell divisions are deterministic and intrinsically regulated in young and old age, they are variable and niche sensitive in mid-life. These results indicate that the balance between intrinsic and extrinsic regulation of stem cell activity alters substantially with age and help explain why stem cell numbers increase through life, yet regenerative potency declines. Changes in stem cell activity may underpin aging. We trained an artificial neural network to interpret gene-expression patterns of paired daughter cells from individual stem cell divisions. Our results show that the self-renewal ability of individual stem cells alters substantially with age and help explain why stem cell numbers increase through life, yet regenerative potency declines.