Host &ndash Pathogen Protein Interactions Predicted by Structure Introduction | Method | Predictions

The protocol begins with the target set of host and pathogen protein sequences (Fig. 1). These sequences are first modeled by an automated comparative modeling pipeline that detects similarities to all known protein sequences and structures. Pairs of host and pathogen proteins that each have detectable similarity to components of a known interaction are listed as putative interactions. If the known interaction has an experimentally determined structure, models are built for the host and pathogen target sequences, and their putative interface is assessed using a statistical potential score. In the absence of structure, the significance of the sequence similarity for the host and pathogen proteins to their interaction templates is assessed. These potential interactions are then filtered by the biological context of their component proteins, such as life-cycle stage or tissue expression.

Fig 1. Prediction Protocol

questions: davisf at janelia dot hhmi dot org