Proper control of gene expression is governed by the complex interplay of many transcription factors. One major component governing where and how a transcription factor binds to genomic DNA is its inherent sequence binding preferences. In our lab, we evaluate models of transcription factor binding preferences, with the end goal of accurate prediction of transcription factor occupancy in vivo. In addition to accurate sequence preference models, we develop methods integrating multiple sources of information, including DNA accessibility, protein interactions, DNA looping, epigenetic modifications, and gene expression.