Computational protocol: The multiple-specificity landscape of modular peptide recognition domains

Similar protocols

Protocol publication

[…] The experimental data sets used in this work come from large-scale phage display experiments (). For PDZ domains, all phage display peptides found in for wild-type domains in humans and worms were used in our analysis. For each domain, the peptides were aligned from the C terminus. Owing to the presence of STOP codons in the phage library, some peptides are shorter than seven amino acids (missing residues are labeled with X). For yeast SH3 domains, the phage peptides from were automatically aligned with the MUSCLE alignment software () using settings that prevented internal gaps. The new data for the three WW domains come from a recent phage display experiment run with similar protocols as for the PDZ and SH3 domains (see for the raw data). Because of the amplification step in phage display, the frequency of the peptides pulled out experimentally is difficult to interpret in terms of binding strength. For this reason, peptides retrieved multiple times were treated as unique throughout our analysis. [...] The Rosetta software () was used to dock the ligand EETDIW to DLG1 PDZ1, using the recent PDB structure 2WL7 for the PDZ domain. Ligand position was optimized with Rosetta2.3 allowing for backbone flexibility, and the highest scoring trajectory out of 100 optimization runs was used in our model. Binding energies and residue preferences in the peptides were analyzed with FoldX (; see ). All structures were visualized with Pymol (http://www.pymol.org). […]

Pipeline specifications

Software tools MUSCLE, FoldX, PyMOL
Applications Protein structure analysis, Nucleotide sequence alignment
Organisms Homo sapiens