The collision of 'big computing' ' petascale systems and the multicore revolution ' and 'big biology' ' high-throughput sequencing and the '-omics' revolution ' have ushered the development of new computational paradigms to store and process voluminous experimental and simulation data. Analysis and interpretation of these data present new technical challenges, but exciting opportunities to advance our biological understanding. I discuss two applications of machine learning techniques in biology: determination of the folding pathways of an antimicrobial 'lasso' peptide, and inference of HIV 'fitness landscapes' for rational vaccine design.