Researchers from China and the US introduce an online tool for taking a look at transcription factor-DNA binding intensities in different cell types, along with the variants that affect them. The approach — known as "deep learning-based functional impact of non-coding variants evaluator," or DeFine — uses deep learning models to tease out transcription factor binding intensities and non-coding SNP impacts on these binding events based on chromatin immunoprecipitation sequencing data.