A major objective of modern biology is to decipher the mechanistic principles underlying phenotypic variation. It is now possible to profile the molecular state of millions of individual cells at once using a powerful technique called single-cell RNA sequencing (scRNA-seq). As scRNA-seq datasets are increasingly generated from different organisms, comparative analysis can help identify similarities and differences at the molecular level that shape phenotypic diversity. I will introduce how we can utilise supervised analysis to identify evolutionary correspondences across species. This approach paves the way to substantially enhance our understanding of how molecular changes influence cellular evolution. I will also introduce how we are using this framework to learn about disease causing mutations in our lab.