Computational psychiatry: a report from the 2017 NIMH workshop on opportunities and challenges

Psychiatry is at a crossroads. Neuroscience has made substantial progress toward understanding the link between brain mechanisms and behavior, but incorporating this understanding into psychiatric practice has proven challenging. One potential way forward is to use computational techniques [ 1 , 2 ]. Computational techniques can be harnessed to develop models that clearly link neural systems to behavior [ 3 , 4 ]), allowing us to develop concrete theoretical conceptualizations that bidirectionally link levels of analysis, from molecules to circuits and from circuits to behavior [ 3 , 5 , 6 ]. In addition, to identify regularities and other consistent, identifiable features in large psychiatric data sets, computational techniques may enable identification, categorization, and prediction of dimensional processes in heterogeneous psychiatric populations [ 3 , 7 ]. To determine how computational techniques and perspectives could inform psychiatric practice and how psychiatric studies could drive new directions in computational neuroscience, the National Institute of Mental Health (NIMH) organized a 2-day meeting (26–27 June 2017) to bring together leaders in experimental and computational neuroscience and clinical and translational psychiatry to discuss and define the opportunities and challenges that exist in the field of Computational Psychiatry (see the new Computational Psychiatry Program and the Theoretical and Computational Neuroscience Program at NIMH for related funding opportunities). Approximately 90 investigators and NIMH staff attended the workshop. The list of participants can be found here and detailed minutes of the meeting are available upon request.