Skip to main content
Fig. 5 | BMC Bioinformatics

Fig. 5

From: rstoolbox - a Python library for large-scale analysis of computational protein design data and structural bioinformatics

Fig. 5

Comparison and benchmarking of different design protocols. Representation of the results obtained using four different design protocols. With the exception of the panel identifiers, the image was directly created with the code presented in Table 4. a Representation of four scoring metrics in the design of a new protein binder. score – shows the overall Rosetta score; RMSD – root mean square deviation relative to BINDI; ddG –Rosetta energy for the interaction between two proteins; bb_clash - quantifies the backbone clashes between the binder and the target protein; b BLOSUM62 positional sequence score for the top design of the no_target (blue) and pack (green) design populations showcases how to analyse and compare individual decoys. The higher the value, the more likely two residue types (design vs. BINDI) are to interchange within evolutionary related proteins. Special regions of interest can be easily highlighted, as for instance the binding region (highlighted in salmon). c Population-wide analysis of the sequence recovery of the binding motif region for no_target and pack simulations. Darker shades of blue indicate a higher frequency and green frames indicate the reference residue type (BINDI sequence). This representation shows that the pack population explores more frequently residue types found in the BINDI design in the region of the binding motif

Back to article page