Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Jul;41(Web Server issue):W8-14.
doi: 10.1093/nar/gkt427. Epub 2013 May 28.

MISTIC: Mutual Information Server to Infer Coevolution

Affiliations
Free PMC article

MISTIC: Mutual Information Server to Infer Coevolution

Franco L Simonetti et al. Nucleic Acids Res. .
Free PMC article

Abstract

MISTIC (mutual information server to infer coevolution) is a web server for graphical representation of the information contained within a MSA (multiple sequence alignment) and a complete analysis tool for Mutual Information networks in protein families. The server outputs a graphical visualization of several information-related quantities using a circos representation. This provides an integrated view of the MSA in terms of (i) the mutual information (MI) between residue pairs, (ii) sequence conservation and (iii) the residue cumulative and proximity MI scores. Further, an interactive interface to explore and characterize the MI network is provided. Several tools are offered for selecting subsets of nodes from the network for visualization. Node coloring can be set to match different attributes, such as conservation, cumulative MI, proximity MI and secondary structure. Finally, a zip file containing all results can be downloaded. The server is available at http://mistic.leloir.org.ar. In summary, MISTIC allows for a comprehensive, compact, visually rich view of the information contained within an MSA in a manner unique to any other publicly available web server. In particular, the use of circos representation of MI networks and the visualization of the cumulative MI and proximity MI concepts is novel.

Figures

Figure 1.
Figure 1.
Circos representation of the Thioredoxin protein family (PF00085). Labels in the first (outer) circle indicate the amino acid code and the PDB number of the reference sequence. Colored square boxes of the second circle indicate the KL conservation score (from red to cyan, red: highest; cyan: lowest). The third and fourth circles show the cMI and pMI scores as histograms, facing outward and inward, respectively. Lines in the center of the circle connect pairs of positions with MI score >6.5. Red lines represent the top 5%; black ones are between 70 and 95%, while gray ones account for the last 70%. Orange bars indicate the three regions mentioned in the text.
Figure 2.
Figure 2.
The MI network and distance network of the Thioredoxin protein family (PF00085). Panel (A) Mutual information network (PF00085). Panel (B) Distance network of the reference structure (PDB code: 2trx). Amino acids are represented as circles colored from red to cyan on conservation (from higher to lower). Edges are represented as lines binding nodes if they have a significant MI value (MI score > 6.5) or are closer than 5 Å (panels A and B, respectively).
Figure 3.
Figure 3.
MISTIC interactive framework showing in panel (A) part of the PF00085 MI network with several nodes and edges selected (selected circles are highlighted in red and selected edges in yellow). Panel (B) shows the amino acid frequency when only one node is selected (e.x: 75). Panel (C) shows the KL sequence logo of the (several) selected nodes.
Figure 4.
Figure 4.
Left: Mutual information network (PF00085) colored from red to cyan on conservation (from higher to lower). Right: ribbon representation of the reference structure (PDB code: 2trx) colored as the network style (conservation). Also, catalytic residues are highlighted with Cα VDW style.
Figure 5.
Figure 5.
(A) MI network colored on residue cMI scores from violet to yellow (higher to lower). Residues with highest cMI were selected by filtering the network (upon cMI). (B): ribbon representation of the reference structure (PDB code: 2trx) colored as the network style (cMI score). Selected subnetwork as well as the catalytic residues are highlighted with VDW style.

Similar articles

See all similar articles

Cited by 55 articles

See all "Cited by" articles

References

    1. Gloor GB, Martin LC, Wahl LM, Dunn SD. Mutual information in protein multiple sequence alignments reveals two classes of coevolving positions. Biochemistry. 2005;44:7156–7165. - PubMed
    1. Martin LC, Gloor GB, Dunn SD, Wahl LM. Using information theory to search for co-evolving residues in proteins. Bioinformatics. 2005;21:4116–4124. - PubMed
    1. Tillier ER, Lui TW. Using multiple interdependency to separate functional from phylogenetic correlations in protein alignments. Bioinformatics. 2003;19:750–755. - PubMed
    1. Atchley WR, Wollenberg KR, Fitch WM, Terhalle W, Dress AW. Correlations among amino acid sites in bHLH protein domains: an information theoretic analysis. Mol. Biol. Evol. 2000;17:164–178. - PubMed
    1. Yip K, Patel P, Kim PM, Engelman DM, McDermott D, Gerstein M. An integrated system for studying residue coevolution in proteins. Bioinformatics. 2008;24:290–292. - PubMed

Publication types

Feedback