Machine learning techniques are having a huge impact on how biologists study and understand proteins. Protein structure prediction has been revolutionized, and new tools are improving functional annotation of proteins, as well as opening up new possibilities for protein design. Written and edited by experts in the field, this collection from Cold Spring Harbor Perspectives in Biology explores the rapidly evolving intersection of machine learning and protein science. The contributors review various approaches for learning representations of proteins, as well as statistical models of co-evolution and large-scale homology searches, which have important implications for protein structure prediction. In addition, they examine applications of machine learning for functional annotation of proteins and variant effect prediction. The collection also explores generative models for protein sequence and structure and looks at the environmental impact of applying these tools, acknowledging the