Application of Mutual Information measures to find covarying positions between proteins in viral families

JAVIER A. ISERTE; FRANCO SIMONETTI; CRISTINA MARINO BUSLJE

Rosario, Santa Fe

Congreso; 4to. Congreso Argentino de Bioinformática y Biología Computacional (4CAB2C) y 4ta. Conferencia Internacional de la Sociedad Iberoamericana de Bioinformática (SolBio); 2013

Asociación Argentina de Bioinformática y Biología Computacional.

BackgroundMutual Information (MI) is a method used for detecting covarying positions in protein families. Its utilization canbe impaired by several drawbacks, mainly: requirement of large numbers of sequences and signal noise imposed byphylogenetic relationships. Several modified measures based on MI were developed in the last years to overcomethese difficulties [1?4]. MI measures are commonly applied to detect covarying residues in a single protein (intraprotein), but it can also be applied to detect covariation between residues of different proteins (inter proteins). Themajor difficulty for the detection of covariation between proteins is to obtain a large set of protein sequences forwhich we know their evolutionary relationship (organism co-localization and orthology). That can be overcomederiving the sequences from full genomes, making viruses a suitable and unexplored candidate model to studycoevolution between proteins. Detection of covarying residues between proteins could help to predict proteininteractions and provide insights into important biological mechanisms. Also, this could be used to detect importantfunctional residues not conserved in viral proteins that are undetected by other methods.Materials And MethodsTaxonomy for Viral families was obtained from ICTVdb. Viral sequences were collected from the generalpurpose Nucleotide database from NCBI, and others curated, family or species specific databases.[HIV SequenceDatabase(http://www.hiv.lanl.gov), Virus Variation (http://www.ncbi.nlm.nih.gov /genomes/VirusVariation/),HVDB (http://s2as02.genes.nig.ac.jp), DengueDB (http://www.denguedb.org), Viral Sequece Database(http://kcdc.labkm.net/vsd/database)] Sequence clustering was performed using Usearch software and Hobohm-1method and sequence alignments were built using Muscle. MI calculations were performed using MISTIC webserver ( http://mistic.leloir.org.ar )  [5].ResultsWe obtained the viral genomic sequences for 96 viral families. Different sequences clustering (according tosequence identity) were analyzed in the study. Families showing low number of sequence clusters were discardedfor subsequent analysis. This led us to obtain a set of 28 viral families. In this work, we show the results ofcovarying residues between proteins of three selected families containing pathogens of high impact in publichealth: Flaviviridae, Retroviridae, and Picornaviridae.ConclusionsDifferent conditions were explored to detect coevolution between proteins in viral models. Our results provideknowledge regarding coevolving residues pairs and networks in viral pathogens. That could shed light on theimplications of particular proteins or group of residues in a functional role. Also, this kind of analysis helpsproviding a rational frame for experiments design.References1. Dunn SD, Wahl LM, Gloor GB: Mutual information without the influence of phylogeny or entropy dramatically improves residue contact prediction. Bioinformatics 2008, 24 (3):333?340.2. White RA, Szurmant H, Hoch JA, Hwa T: Features of protein-protein interactions in two-component signaling deduced from genomic libraries. Meth. Enzymol. 2007, 422:75?101.3. Weigt M, White RA, Szurmant H, Hoch JA, Hwa T: Identification of direct residue contacts in protein-protein interaction by message passing. Proc. Natl. Acad. Sci. U.S.A. 2009, 106:67?72.4. Buslje CM, Santos J, Delfino JM, Nielsen M:Correction for phylogeny, small number of observations and data redundancy improves the identification of coevolving amino acid pairs using mutual information. Bioinformatics 2009, 25(9):1125?1131.5. Simonetti FL, Teppa E, Chernomoretz A, Nielsen M, Marino Buslje C: MISTIC: mutual information server to infer coevolution.Nucleic Acids Res. 2013, 41(W1):W8?W14.