Our research focuses on understanding how the brain encodes and processes somatosensory stimuli. Somatosensory stimuli, such as touch, temperature and pain, inform animals about innocuous and harmful stimuli in their immediate environment. We use larval zebrafish as a model to map out somatosensory circuits, combining neural activity recording with behavioral observations. Using machine learning we use such data to develop realistic circuit models that quantitatively describe how information is processed by the brain. We use these models to suggest hypotheses for targeted experiments that analyze circuit anatomy as well as cellular mechanisms of neural processing, with the ultimate goal to gain a full mechanistic understanding of vertebrate somatosensation.