With great technological advances and numerous ‘big data’ initiatives targeted at generating and acquiring large amounts of biomedical information, there has been an astonishing growth in the volume of data in recent years. Considering sequencing alone, the size of data has approximately doubled every six months in the last decade. Continuing at this rate, we can expect to reach a zettabyte of sequencing data generated per year by 2025. In addition to the challenges associated with ever increasing data size, biomedical data is often multidimensional, i.e. it may include clinical measurements from electronic health records (EHRs), drug usage data, mRNA expression levels, environmental exposures, and others. Furthermore, this data is often incomplete, noisy, and heterogeneous (categorical, continuous, or binary), which introduces significant difficulties in analysis. Data-driven methods, such as pattern recognition, embrace and leverage data complexity through reduction, classification, and clustering to help elucidate features and structures necessary for interpretation. The PSB 2017 session titled “Patterns in Biomedical Data - How do we find them?” is dedicated to the presentation of innovative, data-driven pattern recognition methods and their applications to biomedical research and precision medicine. We encourage submissions of novel pattern recognition approaches, applications of existing methods to a wide range of biomedical data types, as well as manuscripts on combating current challenges in big data analysis and interpretation.