Predicting and annotating catalytic residues: an information theoretic approach.

We introduce a computational method to predict and annotate the catalytic residues of a protein using only its sequence information, so that we describe both the residues' sequence locations (prediction) and their specific biochemical roles in the catalyzed reaction (annotation). While knowing the chemistry of an enzyme's catalytic residues is essential to understanding its function, the challenges of prediction and annotation have remained difficult, especially when only the enzyme's sequence and no homologous structures are available. Our sequence-based approach follows the guiding principle that catalytic residues performing the same biochemical function should have similar chemical environments; it detects specific conservation patterns near in sequence to known catalytic residues and accordingly constrains what combination of amino acids can be present near a predicted catalytic residue. We associate with each catalytic residue a short sequence profile and define a Kullback-Leibler (KL) distance measure between these profiles, which, as we show, effectively captures even subtle biochemical variations. We apply the method to the class of glycohydrolase enzymes. This class includes proteins from 96 families with very different sequences and folds, many of which perform important functions. In a cross-validation test, our approach correctly predicts the location of the enzymes' catalytic residues with a sensitivity of 80% at a specificity of 99.4%, and in a separate cross-validation we also correctly annotate the biochemical role of 80% of the catalytic residues. Our results compare favorably to existing methods. Moreover, our method is more broadly applicable because it relies on sequence and not structure information; it may, furthermore, be used in conjunction with structure-based methods.     

The identification of conserved interactions within the SH3 domain by alignment of sequences and structures

The SH3 domain, comprised of approximately 60 residues, is found within a wide variety of proteins, and is a mediator of protein-protein interactions. Due to the large number of SH3 domain sequences and structures in the databases, this domain provides one of the best available systems for the examination of sequence and structural conservation within a protein family. In this study, a large and diverse alignment of SH3 domain sequences was constructed, and the pattern of conservation within this alignment was compared to conserved structural features, as deduced from analysis of eighteen different SH3 domain structures. Seventeen SH3 domain structures solved in the presence of bound peptide were also examined to identify positions that are consistently most important in mediating the peptide-binding function of this domain. Although residues at the two most conserved positions in the alignment are directly involved in peptide binding, residues at most other conserved positions play structural roles, such as stabilizing turns or comprising the hydrophobic core. Surprisingly, several highly conserved side-chain to main-chain hydrogen bonds were observed in the functionally crucial RT-Src loop between residues with little direct involvement in peptide binding. These hydrogen bonds may be important for maintaining this region in the precise conformation necessary for specific peptide recognition. In addition, a previously unrecognized yet highly conserved beta-bulge was identified in the second beta-strand of the domain, which appears to provide a necessary kink in this strand, allowing it to hydrogen bond to both sheets comprising the fold.     

Macromolecular docking restrained by a small angle X-ray scattering profile

While many structures of single protein components are becoming available, structural characterization of their complexes remains challenging. Methods for modeling assembly structures from individual components frequently suffer from large errors, due to protein flexibility and inaccurate scoring functions. However, when additional information is available, it may be possible to reduce the errors and compute near-native complex structures. One such type of information is a small angle X-ray scattering (SAXS) profile that can be collected in a high-throughput fashion from a small amount of sample in solution. Here, we present an efficient method for protein–protein docking with a SAXS profile (FoXSDock): generation of complex models by rigid global docking with PatchDock, filtering of the models based on the SAXS profile, clustering of the models, and refining the interface by flexible docking with FireDock. FoXSDock is benchmarked on 124 protein complexes with simulated SAXS profiles, as well as on 6 complexes with experimentally determined SAXS profiles. When induced fit is less than 1.5 Å interface C_α RMSD and the fraction residues of missing from the component structures is less than 3%, FoXSDock can find a model close to the native structure within the top 10 predictions in 77% of the cases; in comparison, docking alone succeeds in only 34% of the cases. Thus, the integrative approach significantly improves on molecular docking alone. The improvement arises from an increased resolution of rigid docking sampling and more accurate scoring.     

An integrated approach using orthogonal analytical techniques to characterize heparan sulfate structure

Heparan sulfate (HS), a glycosaminoglycan present on the surface of cells, has been postulated to have important roles in driving both normal and pathological physiologies. The chemical structure and sulfation pattern (domain structure) of HS is believed to determine its biological function, to vary across tissue types, and to be modified in the context of disease. Characterization of HS requires isolation and purification of cell surface HS as a complex mixture. This process may introduce additional chemical modification of the native residues. In this study, we describe an approach towards thorough characterization of bovine kidney heparan sulfate (BKHS) that utilizes a variety of orthogonal analytical techniques (e.g. NMR, IP-RPHPLC, LC-MS). These techniques are applied to characterize this mixture at various levels including composition, fragment level, and overall chain properties. The combination of these techniques in many instances provides orthogonal views into the fine structure of HS, and in other instances provides overlapping / confirmatory information from different perspectives. Specifically, this approach enables quantitative determination of natural and modified saccharide residues in the HS chains, and identifies unusual structures. Analysis of partially digested HS chains allows for a better understanding of the domain structures within this mixture, and yields specific insights into the non-reducing end and reducing end structures of the chains. This approach outlines a useful framework that can be applied to elucidate HS structure and thereby provides means to advance understanding of its biological role and potential involvement in disease progression. In addition, the techniques described here can be applied to characterization of heparin from different sources.     

Structural clusters of evolutionary trace residues are statistically significant and common in proteins

Given the massive increase in the number of new sequences and structures, a critical problem is how to integrate these raw data into meaningful biological information. One approach, the Evolutionary Trace, or ET, uses phylogenetic information to rank the residues in a protein sequence by evolutionary importance and then maps those ranked at the top onto a representative structure. If these residues form structural clusters, they can identify functional surfaces such as those involved in molecular recognition. Now that a number of examples have shown that ET can identify binding sites and focus mutational studies on their relevant functional determinants, we ask whether the method can be improved so as to be applicable on a large scale. To address this question, we introduce a new treatment of gaps resulting from insertions and deletions, which streamlines the selection of sequences used as input. We also introduce objective statistics to assess the significance of the total number of clusters and of the size of the largest one. As a result of the novel treatment of gaps, ET performance improves measurably. We find evolutionarily privileged clusters that are significant at the 5% level in 45 out of 46 (98%) proteins drawn from a variety of structural classes and biological functions. In 37 of the 38 proteins for which a protein-ligand complex is available, the dominant cluster contacts the ligand. We conclude that spatial clustering of evolutionarily important residues is a general phenomenon, consistent with the cooperative nature of residues that determine structure and function. In practice, these results suggest that ET can be applied on a large scale to identify functional sites in a significant fraction of the structures in the protein databank (PDB). This approach to combining raw sequences and structure to obtain detailed insights into the molecular basis of function should prove valuable in the context of the Structural Genomics Initiative.     

Prediction and confirmation of a switch-like region within the N-terminal domain of hSIRT1

Many proteins display conformational changes resulting from allosteric regulation. Often only a few residues are crucial in conveying these structural and functional allosteric changes. These regions that undergo a significant change in structure upon receiving an input signal, such as molecular recognition, are defined as switch-like regions. Identifying these key residues within switch-like regions can help elucidate the mechanism of allosteric regulation and provide guidance for synthetic regulation. In this study, we combine a novel computational workflow with biochemical methods to identify a switch-like region in the N-terminal domain of human SIRT1 (hSIRT1), a lysine deacetylase that plays important roles in regulating cellular pathways. Based on primary sequence, computational methods predicted a region between residues 186–193 in hSIRT1 to exhibit switch-like behavior. Mutations were then introduced in this region and the resulting mutants were tested for allosteric reactions to resveratrol, a known hSIRT1 allosteric regulator. After fine-tuning the mutations based on comparison of known secondary structures, we were able to pinpoint M193 as the residue essential for allosteric regulation, likely by communicating the allosteric signal. Mutation of this residue maintained enzyme activity but abolished allosteric regulation by resveratrol. Our findings suggest a method to predict switch-like regions in allosterically regulated enzymes based on the primary sequence. If further validated, this could be an efficient way to identify key residues in enzymes for therapeutic drug targeting and other applications.     

Profiling gene expression using onto-express

Gene expression profiles obtained through microarray or data mining analyses often exist as vast data strings. To interpret the biology of these genetic profiles, investigators must analyze this data in the context of other information such as the biological, biochemical, or molecular function of the translated proteins. This is particularly challenging for a human analyst because large quantities of less than relevant data often bury such information. To address this need we implemented an automated routine, called Onto-Express (http://vortex.cs.wayne.edu:8080), to systematically translate genetic fingerprints into functional profiles. Using strings of accession or cluster identification numbers, Onto-Express searches the public databases and returns tables that correlate expression profiles with the cytogenetic locations, biochemical and molecular functions, biological processes, cellular components, and cellular roles of the translated proteins. The profiles created by Onto-Express fundamentally increase the value of gene expression analyses by facilitating the translation of quantitative value sets to records that contain biological implications.     

Biological roles and functional mechanisms of arenavirus z protein in viral replication

Arenaviruses can cause severe hemorrhagic fever diseases in humans, with limited prophylactic or therapeutic measures. A small RING-domain viral protein Z has been shown to mediate the formation of virus-like particles and to inhibit viral RNA synthesis, although its biological roles in an infectious viral life cycle have not been directly addressed. By taking advantage of the available reverse genetics system for a model arenavirus, Pichinde virus (PICV), we provide the direct evidence for the essential biological roles of the Z protein's conserved residues, including the G2 myristylation site, the conserved C and H residues of RING domain, and the poorly characterized C-terminal L79 and P80 residues. Dicodon substitutions within the late (L) domain (PSAPPYEP) of the PICV Z protein, although producing viable mutant viruses, have significantly reduced virus growth, a finding suggestive of an important role for the intact L domain in viral replication. Further structure-function analyses of both PICV and Lassa fever virus Z proteins suggest that arenavirus Z proteins have similar molecular mechanisms in mediating their multiple functions, with some interesting variations, such as the role of the G2 residue in blocking viral RNA synthesis. In summary, our studies have characterized the biological roles of the Z protein in an infectious arenavirus system and have shed important light on the distinct functions of its domains in virus budding and viral RNA regulation, the knowledge of which may lead to the development of novel antiviral drugs.     

Integration of metabolomics and proteomics in molecular plant physiology--coping with the complexity by data-dimensionality reduction

In recent years, genomics has been extended to functional genomics. Toward the characterization of organisms or species on the genome level, changes on the metabolite and protein level have been shown to be essential to assign functions to genes and to describe the dynamic molecular phenotype. Gas chromatography (GC) and liquid chromatography coupled to mass spectrometry (GC- and LC-MS) are well suited for the fast and comprehensive analysis of ultracomplex metabolite samples. For the integration of metabolite profiles with quantitative protein profiles, a high throughput (HTP) shotgun proteomics approach using LC-MS and label-free quantification of unique proteins in a complex protein digest is described. Multivariate statistics are applied to examine sample pattern recognition based on data-dimensionality reduction and biomarker identification in plant systems biology. The integration of the data reveal multiple correlative biomarkers providing evidence for an increase of information in such holistic approaches. With computational simulation of metabolic networks and experimental measurements, it can be shown that biochemical regulation is reflected by metabolite network dynamics measured in a metabolomics approach. Examples in molecular plant physiology are presented to substantiate the integrative approach.     

Imprint of evolutionary conservation and protein structure variation on the binding function of protein tyrosine kinases

MOTIVATION: According to the models of divergent molecular evolution, the evolvability of new protein function may depend on the induction of new phenotypic traits by a small number of mutations of the binding site residues. Evolutionary relationships between protein kinases are often employed to infer inhibitor binding profiles from sequence analysis. However, protein kinases binding profiles may display inhibitor selectivity within a given kinase subfamily, while exhibiting cross-activity between kinases that are phylogenetically remote from the prime target. The emerging insights into kinase function and evolution combined with a rapidly growing number of publically available crystal structures of protein kinases complexes have motivated structural bioinformatics analysis of sequence-structure relationships in determining the binding function of protein tyrosine kinases. RESULTS: In silico profiling of Imatinib mesylate and PD-173955 kinase inhibitors with protein tyrosine kinases is conducted on kinome scale by using evolutionary analysis and fingerprinting inhibitor-protein interactions with the panel of all publically available protein tyrosine kinases crystal structures. We have found that sequence plasticity of the binding site residues alone may not be sufficient to enable protein tyrosine kinases to readily evolve novel binding activities with inhibitors. While evolutionary signal derived solely from the tyrosine kinase sequence conservation can not be readily translated into the ligand binding phenotype, the proposed structural bioinformatics analysis can discriminate a functionally relevant kinase binding signal from a simple phylogenetic relationship. The results of this work reveal that protein conformational diversity is intimately linked with sequence plasticity of the binding site residues in achieving functional adaptability of protein kinases towards specific drug binding. This study offers a plausible molecular rationale to the experimental binding profiles of the studied kinase inhibitors and provides a theoretical basis for constructing functionally relevant kinase binding trees.     

Homology-based modeling of 3D structures of protein-protein complexes using alignments of modified sequence profiles

Customary practice in predicting 3D structures of protein-protein complexes is employment of various docking methods when the structures of separate monomers are known a priori. The alternative approach, i.e. the template-based prediction with pure sequence information as a starting point, is still considered as being inferior mostly due to presumption that the pool of available structures of protein-protein complexes, which can serve as putative templates, is not sufficiently large. Recently, however, several labs have developed databases containing thousands of 3D structures of protein-protein complexes, which enable statistically reliable testing of homology-based algorithms. In this paper we report the results on homology-based modeling of 3D structures of protein complexes using alignments of modified sequence profiles. The method, called HOMology-BAsed COmplex Prediction (HOMBACOP), has two distinctive features: (I) extra weight on aligning interfacial residues in the dynamical programming algorithm, and (II) increased gap penalties for the interfacial segments. The method was tested against our recently developed ProtCom database and against the Boston University protein-protein BENCHMARK. In both cases, models generated were compared to the models built on basis of customarily protein structure initiative (PSI)-BLAST sequence alignments. It was found that existence of homologous (by the means of PSI-BLAST) templates (44% of cases) enables both methods to produce models of good quality, with the profiles method outperforming the PSI-BLAST models (with respect to the percentage of correctly predicted residues on the complex interface and fraction of native interfacial contacts). The models were evaluated according to the CAPRI assessment criteria and about two thirds of the models were found to fall into acceptable and medium-quality categories. The same comparison of a larger set of 463 protein complexes showed again that profiles generate better models. We further demonstrate, using our ProtCom database, the suitability of the profile alignment algorithm in detecting remote homologues between query and template sequences, where the PSI-BLAST method fails.     

Comprehensive modeling and functional analysis of Toll‐like receptor ligand‐recognition domains

Toll‐like receptors (TLRs) are innate immune pattern‐recognition receptors endowed with the capacity to detect microbial pathogens based on pathogen‐associated molecular patterns. The understanding of the molecular principles of ligand recognition by TLRs has been greatly accelerated by recent structural information, in particular the crystal structures of leucine‐rich repeat‐containing ectodomains of TLR2, 3, and 4 in complex with their cognate ligands. Unfortunately, for other family members such as TLR7, 8, and 9, no experimental structural information is currently available. Methods such as X‐ray crystallography or nuclear magnetic resonance are not applicable to all proteins. Homology modeling in combination with molecular dynamics may provide a straightforward yet powerful alternative to obtain structural information in the absence of experimental (structural) data, provided that the generated three‐dimensional models adequately approximate what is found in nature. Here, we report the development of modeling procedures tailored to the structural analysis of the extracellular domains of TLRs. We comprehensively compared secondary structure, torsion angles, accessibility for glycosylation, surface charge, and solvent accessibility between published crystal structures and independently built TLR2, 3, and 4 homology models. Finding that models and crystal structures were in good agreement, we extended our modeling approach to the remaining members of the TLR family from human and mouse, including TLR7, 8, and 9.     

Structures of thermophilic and mesophilic adenylate kinases from the genus Methanococcus

The crystal structures of adenylate kinases from the thermophile Methanococcus thermolithotrophicus and the mesophile Methanococcus voltae have been solved to resolutions of 2.8A and 2.5A, respectively. The structures of the enzymes are similar to that of the adenylate kinase from archaeal Sulfolobus acidocaldarius in many respects such as the extended central beta-sheets, the short LID domain, and the trimeric state. The analysis of unligated and AMP-bound subunits of M.voltae suggests that movements of two mobile domains are not independent of each other. The methanococcal structures are examined with respect to their lack of the "invariant" Lys residue within the phosphate-binding loop, and two Arg residues in the LID domain are proposed as substituting residues based on their conservation among archaeal adenylate kinases and mobility within the structures. Since S.acidocaldarius adenylate kinase has the invariant Lys residue as well as the two Arg residues, its phosphate-binding loop is examined and compared with those of other adenylate kinases. On the basis of the comparison and other available biochemical data, the unusual conformation of the Lys residue in S.acidocaldarius adenylate kinase is explained. Despite possessing 78% sequence identity, the methanococcal enzymes exhibit significantly different thermal stabilities. To study the determinants of thermostability, several structural features including salt-links, hydrogen bonds, packing density, surface to volume ratio and buried surface area are compared between the enzymes. From their difference in apolar buried surface area, hydrophobic interaction is proposed to be a basis for the disparate thermostabilities, and the corresponding free energy difference is also estimated. Results of previous mutational studies are interpreted in terms of the crystal structures, and support the importance of hydrophobic interactions in thermostability.     

CancerLectinDB: a database of lectins relevant to cancer

The role of lectins in mediating cancer metastasis, apoptosis as well as various other signaling events has been well established in the past few years. Data on various aspects of the role of lectins in cancer is being accumulated at a rapid pace. The data on lectins available in the literature is so diverse, that it becomes difficult and time-consuming, if not impossible to comprehend the advances in various areas and obtain the maximum benefit. Not only do the lectins vary significantly in their individual functional roles, but they are also diverse in their sequences, structures, binding site architectures, quaternary structures, carbohydrate affinities and specificities as well as their potential applications. An organization of these seemingly independent data into a common framework is essential in order to achieve effective use of all the data towards understanding the roles of different lectins in different aspects of cancer and any resulting applications. An integrated knowledge base (CancerLectinDB) together with appropriate analytical tools has therefore been developed for lectins relevant for any aspect of cancer, by collating and integrating diverse data. This database is unique in terms of providing sequence, structural, and functional annotations for lectins from all known sources in cancer and is expected to be a useful addition to the number of glycan related resources now available to the community. The database has been implemented using MySQL on a Linux platform and web-enabled using Perl-CGI and Java tools. Data for individual lectins pertain to taxonomic, biochemical, domain architecture, molecular sequence and structural details as well as carbohydrate specificities. Extensive links have also been provided for relevant bioinformatics resources and analytical tools. Availability of diverse data integrated into a common framework is expected to be of high value for various studies on lectin cancer biology. CancerLectinDB can be accessed through . Availability: CancerLectinDB is available freely for academic use from , Contact nchandra@serc.iisc.ernet.in for further information.     

New insights into the GABA(A) receptor structure and orthosteric ligand binding: receptor modeling guided by experimental data

GABA(A) receptors (GABA(A)Rs) are ligand gated chloride ion channels that mediate overall inhibitory signaling in the CNS. A detailed understanding of their structure is important to gain insights in, e.g., ligand binding and functional properties of this pharmaceutically important target. Homology modeling is a necessary tool in this regard because experimentally determined structures are lacking. Here we present an exhaustive approach for creating a high quality model of the α(1)β(2)γ(2) subtype of the GABA(A)R ligand binding domain, and we demonstrate its usefulness in understanding details of orthosteric ligand binding. The model was constructed by using multiple templates and by incorporation of knowledge from biochemical/pharmacological experiments. It was validated on the basis of objective energy functions, its ability to account for available residue specific information, and its stability in molecular dynamics (MD) compared with that of the two homologous crystal structures. We then combined the model with extensive structure-activity relationships available from two homologous series of orthosteric GABA(A)R antagonists to create a detailed hypothesis for their binding modes. Excellent agreement with key experimental data was found, including the ability of the model to accommodate and explain a previously developed pharmacophore model. A coupling to agonist binding was thereby established and discussed in relation to activation mechanisms. Our results highlight the importance of critical evaluation and optimization of each step in the homology modeling process. The approach taken here can greatly aid in increasing the understanding of GABA(A)Rs and related receptors where structural insight is limited and reliable models are difficult to obtain.     

Identification of Key Hinge Residues Important for Nucleotide-Dependent Allostery in E. coli Hsp70/DnaK

DnaK is a molecular chaperone that has important roles in protein folding. The hydrolysis of ATP is essential to this activity, and the effects of nucleotides on the structure and function of DnaK have been extensively studied. However, the key residues that govern the conformational motions that define the apo, ATP-bound, and ADP-bound states are not entirely clear. Here, we used molecular dynamics simulations, mutagenesis, and enzymatic assays to explore the molecular basis of this process. Simulations of DnaK''s nucleotide-binding domain (NBD) in the apo, ATP-bound, and ADP/P_i-bound states suggested that each state has a distinct conformation, consistent with available biochemical and structural information. The simulations further suggested that large shearing motions between subdomains I-A and II-A dominated the conversion between these conformations. We found that several evolutionally conserved residues, especially G228 and G229, appeared to function as a hinge for these motions, because they predominantly populated two distinct states depending on whether ATP or ADP/P_i was bound. Consistent with the importance of these “hinge” residues, alanine point mutations caused DnaK to have reduced chaperone activities in vitro and in vivo. Together, these results clarify how sub-domain motions communicate allostery in DnaK.     

Structure-function relationships of glucansucrase and fructansucrase enzymes from lactic acid bacteria.

Lactic acid bacteria (LAB) employ sucrase-type enzymes to convert sucrose into homopolysaccharides consisting of either glucosyl units (glucans) or fructosyl units (fructans). The enzymes involved are labeled glucansucrases (GS) and fructansucrases (FS), respectively. The available molecular, biochemical, and structural information on sucrase genes and enzymes from various LAB and their fructan and alpha-glucan products is reviewed. The GS and FS enzymes are both glycoside hydrolase enzymes that act on the same substrate (sucrose) and catalyze (retaining) transglycosylation reactions that result in polysaccharide formation, but they possess completely different protein structures. GS enzymes (family GH70) are large multidomain proteins that occur exclusively in LAB. Their catalytic domain displays clear secondary-structure similarity with alpha-amylase enzymes (family GH13), with a predicted permuted (beta/alpha)(8) barrel structure for which detailed structural and mechanistic information is available. Emphasis now is on identification of residues and regions important for GS enzyme activity and product specificity (synthesis of alpha-glucans differing in glycosidic linkage type, degree and type of branching, glucan molecular mass, and solubility). FS enzymes (family GH68) occur in both gram-negative and gram-positive bacteria and synthesize beta-fructan polymers with either beta-(2-->6) (inulin) or beta-(2-->1) (levan) glycosidic bonds. Recently, the first high-resolution three-dimensional structures have become available for FS (levansucrase) proteins, revealing a rare five-bladed beta-propeller structure with a deep, negatively charged central pocket. Although these structures have provided detailed mechanistic insights, the structural features in FS enzymes dictating the synthesis of either beta-(2-->6) or beta-(2-->1) linkages, degree and type of branching, and fructan molecular mass remain to be identified.     

Classification of the Environment of Protein Residues

We have studied the classification of the environment of residues within protein structures. Eisenberg's original idea created environmental categories to discriminate between similar residues [Bowie et al., Science (1991), 253, 164–170]. These environments grouped residues based upon their buried surface area, polarity of the surrounding environment, and secondary structure element in which the residue is found. However, Eisenberg's original categories led to incomplete discrimination between residues that only partially substitute for each other. We have expanded on Eisenberg's original idea of environmental categories, by both considering additional contacts in the calculation of the solvent-accessible molecular surface area and by subdividing the environmental plot into regions based upon its theoretical features. Our alternative surface area calculations were used in conjunction with the polarity of the environment of the residue to define a new set of environmental categories. These new categories were able to discriminate between residues such as threonine, valine, and aspartic acid while reflecting the propensity of these residues to substitute for each other.     

The bhopalator: a molecular model of the living cell based on the concepts of conformons and dissipative structures

A molecular model of the living cell has been formulated based on a new theory of enzymic catalysis which takes into account the complementary roles of free energy and genetic information. The elementary units of free energy and genetic information that are necessary and sufficient for effectuating molecular mechanisms responsible for the life of the cell are called conformons. Conformons are visualized as a collection of a small number of catalytic residues of enzymes or segments of nucleic acids that are arranged in space and time with appropriate force vectors so as to cause chemical transformations or physical changes of a substrate or a bound ligand. So defined, conformons provide a plausible molecular means to link the genetic information stored in DNA and its ultimate expression, namely networks of coupled intracellular biochemical reactions and physical processes maintained by a continuous dissipation of free energy--dissipative structures of Prigogine. The proposed model of the living cell appears to possess the potential for bridging the gap between molecular biology and the biology of multicellular systems.     

An integrated probabilistic model for functional prediction of proteins.

We develop an integrated probabilistic model to combine protein physical interactions, genetic interactions, highly correlated gene expression networks, protein complex data, and domain structures of individual proteins to predict protein functions. The model is an extension of our previous model for protein function prediction based on Markovian random field theory. The model is flexible in that other protein pairwise relationship information and features of individual proteins can be easily incorporated. Two features distinguish the integrated approach from other available methods for protein function prediction. One is that the integrated approach uses all available sources of information with different weights for different sources of data. It is a global approach that takes the whole network into consideration. The second feature is that the posterior probability that a protein has the function of interest is assigned. The posterior probability indicates how confident we are about assigning the function to the protein. We apply our integrated approach to predict functions of yeast proteins based upon MIPS protein function classifications and upon the interaction networks based on MIPS physical and genetic interactions, gene expression profiles, tandem affinity purification (TAP) protein complex data, and protein domain information. We study the recall and precision of the integrated approach using different sources of information by the leave-one-out approach. In contrast to using MIPS physical interactions only, the integrated approach combining all of the information increases the recall from 57% to 87% when the precision is set at 57%-an increase of 30%.