Integrating HiTOP and Computational Psychiatry for a New Era of Clinical Science

For decades, advancements in understanding and treating mental illness have been hindered by a categorical psychiatric nosology that fails to describe and explain how symptoms emerge and relate to one another. In response, two separate initiatives have contributed to a renewed sense of optimism for scientific and translational discovery. First, quantitative models of psychopathology improve clinical description by relying on empirical data to provide a more accurate representation of the structure of mental illness. Chief among these approaches is the Hierarchical Taxonomy of Psychopathology (HiTOP), which organizes psychopathology based on patterns of symptom covariation observed across numerous studies. In parallel, computational psychiatry seeks to identify neurocognitive processes that give rise to psychopathology and leverage them to forecast important clinical and functional outcomes. Here, we highlight points of convergence and complementarity between HiTOP and computational psychiatry and propose further integration. HiTOP provides an empirically based approach to understanding the structure of psychopathology, but lacks strong explanations of how symptom dimensions are connected to underlying neurocognitive processes. Conversely, computational psychiatry’s emphasis on cognitive processes and multivariate prediction make it well-suited to linking lower level dynamics with symptom variability. Yet, many contemporary computational psychiatry findings lack specificity and validity as a result of a continued reliance on categorical diagnoses. We conclude by highlighting potential barriers that will need to be addressed to maximize the productivity of future collaborative efforts between these initiatives.