Nature of Protein Family Signatures: Insights from Singular Value Analysis of Position-Specific Scoring Matrices

Abstract

Position-specific scoring matrices (PSSMs) are useful for detecting weak homology in protein sequence analysis, and they are thought to contain some essential signatures of the protein families. In order to elucidate what kind of ingredients constitute such family-specific signatures, we apply singular value decomposition to a set of PSSMs and examine the properties of dominant right and left singular vectors. The first right singular vectors were correlated with various amino acid indices including relative mutability, amino acid composition in protein interior, hydropathy, or turn propensity, depending on proteins. A significant correlation between the first left singular vector and a measure of site conservation was observed. It is shown that the contribution of the first singular component to the PSSMs act to disfavor potentially but falsely functionally important residues at conserved sites. The second right singular vectors were highly correlated with hydrophobicity scales, and the corresponding left singular vectors with contact numbers of protein structures. It is suggested that sequence alignment with a PSSM is essentially equivalent to threading supplemented with functional information. In addition, singular vectors may be useful for analyzing and annotating the characteristics of conserved sites in protein families.

Uncover the conserved property underlying sequence‐distant and structure‐similar proteins

“…Uncovering the conserved properties may shed light on the folding mechanism of proteins and help with the development of computational tools for protein structure prediction 15. Several bioinformatics approaches to this question are carried out through comparative analysis of structural properties,5 analysis of types of amino acid residues in protein sequences,6, 7 and singular value analysis of position‐specific scoring matrices 9. These studies have showed the importance of the conserved hydrophobic or hydropathic profile.…”

“…For example, the known crystal structures of intestinal fatty acid‐binding protein and Manduca sexta fatty acid‐binding protein 2 can be aligned to a 1.62 Å root‐mean‐square deviation (RMSD) of the C α atoms, whereas they share sequence identity of merely 19% 4. These observations raise an interesting question: what properties are actually conserved in the 3D structures of those distant protein sequences?5 Several studies have been performed based on the analysis of amino acid types of individual protein sequences6–8 and using position‐specific scoring matrices 9. It is proposed that the hydrophobic or hydropathic profile is preserved in those similar structures with dissimilar sequences.…”

Cooperative “folding transition” in the sequence space facilitates function-driven evolution of protein families

“…While protein structures can tolerate great many mutations to the extent that proteins with little sequence similarity can share the same fold, residue conservation patterns reflect the structural context of protein sequences. This fact has long been exploited in protein structure Email address: akinjo@protein.osaka-u.ac.jp (Akira R. Kinjo) prediction in the form of position-specific scoring matrices (Taylor, 1986;Gribskov et al, 1987;Altschul et al, 1997;Kinjo and Nakamura, 2008) and, more recently, direct-coupling analysis and related methods (Balakrishnan et al, 2011;Morcos et al, 2011;Jones et al, 2012;Taylor et al, 2012;Miyazawa, 2013;Ekeberg et al, 2013;Kinjo, 2015;Levy et al, 2017). Under a given physiological condition, a polypeptide is either able or unable to fold into some unique structure.…”

A unified statistical model of protein multiple sequence alignment integrating direct coupling and insertions

“…A multiple sequence alignment (MSA) of a family of proteins provides us with valuable information to characterize the protein family in terms of patterns of amino acid residues at alignment sites [ 1 ]. The usefulness of analyzing the residue compositions in the MSA has led to the development of a class of sequence profile methods [ 1 – 3 ] such as PSI-BLAST [ 4 ] and profile hidden Markov models (HMM) [ 5 ], which can be used to detect distantly related proteins, to obtain high-quality alignments, and to improve structure prediction [ 6 ] as well as to characterize functional and structural roles of the conservation pattern [ 7 ]. In the sequence profile methods, it is assumed that the residue composition of each site is independent of other sites.…”