Fragments of local symmetry in a sequence of amino acids: Does one can use for QSPR/QSAR of peptides?

•A model for the physicochemical behaviour of peptides is suggested.•Sequences of amino acids applying as the quasi-SMILES.•The model is based on optimal descriptors calculated with quasi-SMILES.•Measures of symmetry and chaos improve building up the model.•The approach is checking up with several random splits. Most of the protein interactions rely on small domains binding to short peptides. However, neither the number of potential interactions mediated by each domain nor the degree of affinity at a whole proteome level has been studied. Peptide segments involved in 14-3-3 domain-medicated cellular signalling networks were collected from sequence-based datasets of domain-peptide interaction affinities. The affinities peptides measured represented by the Boehringer light units (logBLU) considered as the endpoint for the development of peptide quantitative structure-property/activity relationships (p-QSPR/QSARs). The sequences of amino acids are examined as the structure of the peptide. This approach allows the rational simulate interaction of proteomic systems via amino acid configurations in proteins. Adding the contributions of local symmetry and chaos extracted from the amino acid sequences described here increases the value of the average (over seven different splits into training and control) determination coefficient for the external validation set and reduces its dispersion. [Display omitted]