Prioritization of polygenic combinations for LOAD mouse models from ensemble machine learning on cross‐species transcriptomic signatures

We identified consistent predictions regarding which mouse factors are most influential in modeling disease across model classes and human cohorts. Ensembling the models was a useful approach for feature selection. These priorities informed the selection of mouse strains to breed in the development of polygenic LOAD mouse models by the IU-JAX-PITT MODEL-AD center.