Factor Analysis Scale of Generalized Amino Acid Information as the Source of a New Set of Descriptors for Elucidating the Structure and Activity Relationships of Cationic Antimicrobial Peptides

Abstract

Factor Analysis Scale of Generalized Amino Acid Information (FASGAI), as a new set of amino acid descriptors, reflecting hydrophobicity, alpha and turn propensities, bulky properties, compositional characteristics, local flexibility, and electronic properties, was derived from multi-dimensional properties of 20 coded amino acids. Quantitative Structure -Activity Relationship (QSAR) analyses on 101 cationic Antimicrobial Peptides (AMPs) with 15 residues were then performed using FASGAI representation, Genetic Algorithm (GA) selection features, and Partial Least Squares (PLS) modeling. It was found that electronic properties of the tenth residue, bulky properties of the seventh residue, hydrophobicity of the twelfth residue, electronic properties of the third residue were highly positively correlated to antibacterial activity, whereas, hydrophobicity of the sixth residue, compositional characteristics of the sixth residue, and hydrophobicity of the tenth residue were highly negatively correlated to antibacterial activity. Analysis of variance demonstrated that a specific sequence rather than an overall sequence of AMPs may be responsible for high activity. Favorable results showed that FASGAI descriptors cherish many distinguished advantages such as plentiful structural information, easy manipulation, straightforward physicochemical meaning, and high characterization competence, and GA for parameter selection is valid in QSAR of antimicrobial peptides studied. These have thus pointed us further into the potential direction of investigation on structures and functions of other various functional peptides.

HomoSAR: Bridging comparative protein modeling with quantitative structural activity relationship to design new peptides

“…Shai have proposed a mechanism for the membrane active AMPs which demands an appropriate balance between the hydrophobicity and a net positive charge of the AMPs for the activity on bacteria and this is in agreement with our HomoSAR results. A comparison of the statistics for the HomoSAR model for the CAMEL‐s antibacterial peptides with the QSAR models constructed in previous studies by Stefan, Cherkasov and Jankovic, Liang and Li, and Zhou et al are shown in Table . The statistics for the HomoSAR model in terms of the correlation coefficient ( r 2 ), leave one out cross‐validation ( ) and predictive correlation coefficient ( ) is slightly superior than Liang and Li statistics, whereas the HomoSAR statistics are comparable with that of Zhou et al results.…”

“…A comparison of the statistics for the HomoSAR model for the CAMEL‐s antibacterial peptides with the QSAR models constructed in previous studies by Stefan, Cherkasov and Jankovic, Liang and Li, and Zhou et al are shown in Table . The statistics for the HomoSAR model in terms of the correlation coefficient ( r 2 ), leave one out cross‐validation ( ) and predictive correlation coefficient ( ) is slightly superior than Liang and Li statistics, whereas the HomoSAR statistics are comparable with that of Zhou et al results. The , , , , , , , and statistics also gives credence to the HomoSAR model, with deviation seen in case of the value for .…”

Quantitative structure–activity relationship study of antioxidative peptide by using different sets of amino acids descriptors

“…Based on the distance between atoms and the electro-negativity of each atom, a new set of descriptors call the molecular electro-negativity edge vector (VMEE) be developed by Li et al and be applied to describe the molecular structure of ACE inhibitors and bitter tasting di-peptides [25]. There were reports of the use of other descriptors such as VSTV [26], SZOTT [27], FASGAI [28], and VSW [29]. For the 3D descriptor, in 1995, Collantes et al developed new 3D descriptors for QSAR studies of peptide analogues, named ISA (isotropic surface area)-ECI (electronic charge index) scales with parameterizations that base on three-dimensional structure of amino acid side chains [30].…”

Application of ‘HESH’ Descriptors for the Structure-Activity Relationships of Antimicrobial Peptides

“…Since the pioneering work of peptide QSAR were proposed by Sneath [19] based on physicochemical semi-qualitative data for natural amino acids, a number of quantitative peptide sequence descriptors have been put forward in the past few years [20][21][22][23][24][25][26][27][28][29][30]. Amongst them, the representative ones are listed as follows: Hellberg et al [21] applied orthogonal transformation on 29 physicochemical properties of individual amino acids to produce one set of Z scales (Z); Collantes et al [26] had established two computable 3D descriptors, Isotropic Surface Area (ISA) and Electronic Charge Index (ECI), on the base of three-dimensional structural characters of amino acid side chains.…”

“…The z-scores and ISA-ECI descriptors have proven to be powerful for modeling a variety of biological activities of small peptides with good results. Herein, based on our previous research [27][28][29][30], we used multivariate statistical analyses on 171 amino acid properties to acquire a new set of highly interpretable amino acid descriptors involving hydrophobic properties, steric properties, electronic properties and hydrogen bond contribution.…”