Aptamers are single-stranded oligonucleotides generated by the systematic evolution of ligands by exponential enrichment (SELEX). They have a wide variety of applications, but computational approaches are necessary to efficiently identify candidate aptamers. A team of researchers has now developed RaptGen, a variational autoencoder that efficiently discovers new aptamers not included in the input SELEX dataset. This novel computational model could one day be the standard method for aptamer generation.