Abhishek Tiwari:INTERACTOMICS

From OpenWetWare

(Difference between revisions)
Jump to: navigation, search
m (INTERACTOMICS & BIOLOGICAL NETWORKS)
m (INTERACTOMICS & BIOLOGICAL NETWORKS)
Line 38: Line 38:
'''Synopsis'''
'''Synopsis'''
-
Proteins evolved from a common ancestor are said to be homologues and to constitute a “family” with potentially similar structures, functions, and interactions. The problem of identifying “real” protein families based on amino acid sequence conservation has been the subject of extensive debate, because algorithms that search for pairwise homologies can miss important relations and produce false hits. The availability of a large number of sequenced genomes now allows us to map the full set of protein similarity relationships into a Protein Homology Network (PHN), and protein families appear naturally as dense, highly connected regions of the network. In this study, Medini, Covacci, and Donati describe a new method that identifies these regions of the PHN, and generate a set of protein families (PHN-Families) that correlate with protein function and phylogeny, with a quality comparable to family sets curated by human experts. The method is completely unsupervised and can be applied to any number of genomes. The authors test the biological relevance of the PHN-Families obtained by studying the members of Type III and Type IV secretion systems, showing that this classification can also be used to identify the evolutionary events that led to the formation of multiprotein structures.
+
Proteins evolved from a common ancestor are said to be homologues and to constitute a “family” with potentially similar structures, functions, and interactions. The problem of identifying “real” protein families based on amino acid sequence conservation is still doubtful  because algorithms that search for pairwise homologies can miss important relations and produce false hits. Problem of reconstructing the evolutionary relationships amongst proteins and of classifying them into families from a topological point of view was addressed  by defining the Protein Homology Network (PHN). In the PHN, proteins are seen as nodes connected by links that represent the homology relations inferred by sequence similarity. In such a representation, protein families should appear as dense clusters disconnected from the rest of the network. The availability of a large number of sequenced genomes now allows us to map the full set of protein similarity relationships into a Protein Homology Network (PHN), and protein families appear naturally as dense, highly connected regions of the network. In this study, Medini, Covacci, and Donati describe a new method that identifies these regions of the PHN, and generate a set of protein families (PHN-Families) that correlate with protein function and phylogeny. A comparison with an external protein domain database suggests that this approach produces results with a quality comparable to the ones generated by human experts.The method is completely unsupervised and can be applied to any number of genomes. In summary, the PHN-Families provide a comprehensive catalogue of the protein repertoire, also useful for the detection of inheritance patterns.Given the increasing number of bacterial genome sequences, and the number of genes with unknown function [29,30], PHN-Families could provide a powerful annotation tool, allowing straightforward comparisons of whole genomes and the discovery of novel and previously uncharacterized functions.

Revision as of 07:11, 17 January 2007

Home        Contact        iCODONS        Publications        Research        Projects        Softwares        Links       

INTERACTOMICS & BIOLOGICAL NETWORKS

PLoS Computational Biology Volume 2 | Issue 12 | DECEMBER 2006

  • Modularity and Dynamics of Cellular Networks

Synopsis

At very first sight I can say that Yuan Qi and Hui Ge has written an excellent review article on the recent advances on analyzing the architecture and dynamics of cellular networks. Using mammalian cell signaling as case studies article also summarize how computational modeling yields insight about cell signaling pathways or how computational analyses of networks shed light on specific biological processes. Yuan and Hui also addressed a series of important questions like

1.What are the characteristics of cellular network structures that distinguish them from randomly generated networks?

2.Are the network structures relevant for biological functions? If so, are they evolutionarily conserved and how do they evolve?

3.Are some topological patterns preferred at certain times or conditions?

Modularity and dynamics both underlie the functionality of cellular networks, ranging from transcriptional regulation to cell signaling.Unlike random networks, cellular networks contain characteristic topological patterns that enable their functionality. Modularity exists in a variety of biological contexts, including protein complexes, metabolic pathways, signaling pathways, and transcriptional programs.Network topologies reveal dynamic properties that contribute to cellular functions. Computationally, general graphical models such as dynamic Bayesian networks may be applied to analyze the dynamics of cellular network structures. Article also gives a nice summary of Computational Methods in Network Modeling Using “Omic” Data.


PLoS Computational Biology Volume 2 | Issue 12 | DECEMBER 2006

  • Protein Homology Network Families Reveal Step-Wise Diversification of Type III and Type IV Secretion Systems

Synopsis

Proteins evolved from a common ancestor are said to be homologues and to constitute a “family” with potentially similar structures, functions, and interactions. The problem of identifying “real” protein families based on amino acid sequence conservation is still doubtful because algorithms that search for pairwise homologies can miss important relations and produce false hits. Problem of reconstructing the evolutionary relationships amongst proteins and of classifying them into families from a topological point of view was addressed by defining the Protein Homology Network (PHN). In the PHN, proteins are seen as nodes connected by links that represent the homology relations inferred by sequence similarity. In such a representation, protein families should appear as dense clusters disconnected from the rest of the network. The availability of a large number of sequenced genomes now allows us to map the full set of protein similarity relationships into a Protein Homology Network (PHN), and protein families appear naturally as dense, highly connected regions of the network. In this study, Medini, Covacci, and Donati describe a new method that identifies these regions of the PHN, and generate a set of protein families (PHN-Families) that correlate with protein function and phylogeny. A comparison with an external protein domain database suggests that this approach produces results with a quality comparable to the ones generated by human experts.The method is completely unsupervised and can be applied to any number of genomes. In summary, the PHN-Families provide a comprehensive catalogue of the protein repertoire, also useful for the detection of inheritance patterns.Given the increasing number of bacterial genome sequences, and the number of genes with unknown function [29,30], PHN-Families could provide a powerful annotation tool, allowing straightforward comparisons of whole genomes and the discovery of novel and previously uncharacterized functions.


Oxford Bioinformatics Volume 22 | Number 17 | 1 September 2006

  • Constructing biological networks through combined literature mining and microarray analysis: a LMMA approach

Synopsis

Network representation of biological entities is very important for understanding biological processes, system organization and interaction between entities. Shao Li et al. integrated both the literatures and microarray gene-expression data, and developed a combined literature mining and microarray analysis (LMMA) approach to construct gene networks of a specific biological system. In the LMMA approach, a global network is first constructed using the literature-based co-occurrence method. It is then refined using microarray data.

This work may be a start towards an Integrated Biological Network representation where different data sets from literature, genomics and proteomics will be used to give precise and useful multidimensional interaction networks.

Personal tools