Property distance graph (PD-graph)
The PD-graph calculates the interpeptide PD for a group of sequences and graphically presents their similarity to one another at the PCP level [26]. The program produces a graphic optimizing a metric of how the distances between peptides relate to one another using gradient descent from random initial placement of the sequences. First it automatically calculates the pairwise interpeptide physicochemical property (PCP) similarity (PD, where identical sequences have a PD of 0 and values increase with increasing divergence) of a given set of sequences and then determines through an iterative process the best clustering. PD-graph produces reproducible graphics using a helper script that normalizes the orientation (rotation and reflection) of each graphic to show a consensus of multiple runs. The graph labeling and other features are customized by the user.