Protein Families - Coursework Example

Literature Review Analysing Protein Families with Bioinformatics Programs to Highlight the Importance of Structural and Functional Evolutionary Restraints Literature Review: This thesis deals with the analysis of protein families using bioinformatics programs such as Crescendo/CHARMM/Amber etc. The general protein analysis involves Development of methods to predict structure and function of proteins Studying both the structural and functional components of protein sequences Clustering protein sequences and categorising them with related sequences to develop protein models Grouping similar proteins to generate phylogeny and suggest evolutionary relationships between the proteins studied Analysis of protein structures and functions help in understanding the exact relationships between the proteins and in identifying protein families and features (Gaeta 1997; Lengauer, 2002). The data obtained is organised in such a way so as to access the already existing information on proteins, in this case and develop tools and resources to analyse the data. Development of tools and resources for bioinformatics would involve an integration of computational theory with an understanding of cellular mechanisms. The analysis of proteins can be done at three different levels - 1. at the level of single gene analysis 2. at the level of complete genomes 3. at the level of genes and genomes using functional data. The analysis of a single gene (protein) sequence is done by establishing similarity with other known genes, determining phylogenetic trees and evolutionary relationships; identifying well-defined domains in the sequence, identifying physical properties, binding sites and modification sites comprising of the protein sequence features, predicting sub-cellular localization and predicting secondary and tertiary structures (Gaeta, 1997; Lengauer, 2002). For analysis of complete genomes and to determine which gene families are present in the protein and which are not present, several techniques such as location of genes on the chromosomes, correlation with function or evolution, expansion of duplication of gene families, identification of missing enzymes, and presence or absence of biochemical pathways are considered. Large scale events that have affected evolution of organisms are also studied for genome analysis. When genes or genomes are analysed with respect to functional data the techniques used are expression analysis, micro array data, mRNA concentration measurements, protein concentration measurements, proteomics, and covalent modifications. For analysis of gene and genomes from a functional perspective, comparative analyses of biochemical pathways are made along with deletion or mutant genotypes vs. phenotypes and identification of essential genes, or genes involved in specific processes in structure and functions of the organisms (Lengauer, 2002). Analysis of protein structure and functions are done with the help of computer simulations which are used as tools to investigate protein structure and dynamics under a wide variety of conditions, ranging from ligand binding and enzyme-reaction mechanisms to denaturation (in which protein molecules become biologically inactive) and protein re-folding (Ponder and Case, 2003). In all these simulations, the energy of a protein is represented as a function of its atomic coordinates. These functions are known as force fields as forces on individual atoms are related to the gradient of this potential energy function (Ponder and Case, 2003). The protein force fields are represented using a simple equation that shows the potential energy function (Source, Ponder and Case, 2003) Source, Ponder and Case, 2003 The summations shown here are on bonds, torsions and angles and this equation represents the very basic potential energy function that can reproduce only the essential features of protein energy at an atomic level. According to Ponder and Case (2003), the combination of potential energy function and the other parameters constitute a force field. Protein force fields are thus used to simulate the basic structures and functions of a protein and some of the popular force field programs are CHARMM, Amber and OPLS and GROMOS force fields (Ponder and Case, 2003; Oostenbrink et al, 2005). Some of these force fields and bioinformatics simulation programs are described here to study the relevance of such techniques in protein analysis and its applicability in our present study. Crescendo - This is a recently developed program used for protein-protein docking. The application of this program is based on a functional site prediction method. Crescendo is used to detect protein-protein interaction sites involved in mediating functions and aids in the selection of the correct docking solution from the ones produced using various physical and chemical parameters. Thus crescendo analyzes the degree of variability by checking for unusual substitution patterns using predicted environment, specific substitution tables, and calculates differences. Crescendo helps in identifying functional sites and uses the correct docking solution to detect protein-protein interaction. The aim of this thesis being identification of structural and functional evolutionary constraints, the use of Crescendo seems to be a starting point in our analysis. Amber also known as Assisted Model Building with Energy Refinement was one of the oldest programs in protein simulation and analysis, the Amber molecular mechanics package was developed in the 1980s. For this program, electrostatic potential charges or ESPs were determined and the liquid state simulations or molecular simulations of the liquid state were pioneered at this time. Due to the limited power of computers only polar hydrogens were explicitly represented in these programs. Force constant, idealised bond lengths and angles from crystal structures were taken to match attributes of peptide fragments. The torsional barriers that were extracted had significant electrostatic and van der Waals interactions between the atoms. Thus, Amber helps in refining protein structure and refers to both - a set of molecular mechanical force fields for the simulation of bio molecules used in simulation programs; and a package of molecular simulation programs which includes source code and demos in which the experienced users refines programs to suit their needs. The current version of the code used is Amber version 8. Ponder and Case (2003) give the charge models for Amber potentials. Amber could be used for simulation of bio-molecules and is especially useful for structural analysis of evolutionary constraints. Source, Ponder and Case, 2003 CHARMM or the acronym for Chemistry at HARvard Macromolecular Mechanics is a program for macromolecular simulations, including energy minimization, molecular dynamics and Monte Carlo simulations. Monte Carlo simulation uses random and stochastic methods of probability statistics to select multiple scenarios or simulation possibilities. The current released version of CHARMM is version c32b1. The CHARMM program was also originally developed in the 1980s and initially used the extended atom force fields with no explicit hydrogen atoms. However by 1985 hydrogen atoms bonded to nitrogen and oxygen were explicitly represented. For this a series of super molecular calculations were done for a model compound formamide or N-methylacetamide and the aim was to obtain a balanced interaction between solute-water and water-water energies. CHARMM19 was specifically developed and tested on gas phase simulations although there was recognition for refining parameters to obtain a balance of interaction energies in explicit solvent simulations and resulted in CHARMM22 model. Ponder and Case, 2003 provide the charge models for CHARMM force fields parameterizations. CHARMM being a tool for macromolecular and stochastic Monte Carlo simulations highlights the various molecular interactional patterns and are useful in identifying 'potential' interaction sites. Source, Ponder and Case, 2003 The OPLS or Optimized Potentials for Liquid Simulations placed a strong emphasis on deriving non bonded interactions by comparison to liquid-state thermodynamics (Jorgensen, 1998 as cited in Ponder and Case 2003). The early models of OPLS treated hydrogens bonded to aliphatic carbons as part of an extended atom but represented all other hydrogens explicitly. The initial applications of the OPLS used polar hydrogen atoms only although an all-atom version was developed later. In this case, the parameters were principally derived with reference to condensed phase simulations, although comparison to gas peptide energetics could also be done using this application. The figure below represents the charge models for OPLS force fields. The OPLS-UA represents the early models and the OPLS-AA represents the later all-atom version. Source, Ponder and Case, 2003 Using the foundational concepts of the bioinformatics programs the research objectives in this thesis are geared towards the analysis of protein structures and functions using simulation techniques to determine the propensity for the presence of unusual residues in proteins analysed. The structural and functional evolutionary restraints are also identified using bioinformatics programs to show intermolecular interactional patterns. In a recent study Chelliah et al (2004) use structural information of proteins to identify evolutionary restraints that arise from the structure of the protein and differentiate these restraints from others that derive from intermolecular interactions that mediate functions in the whole organism. Thus restraints that arise from structure of proteins and restraints that arise from intermolecular interactions are differentiated to understand the structural and functional components of protein sequence and the exact role of interactional sites. Chelliah et al point out that gene sequences have helped to identify the structures of proteins although computational methods showing intermolecular interactions can help provide information on functions as well. There is a need to understand the differences between the evolutionary restraints due to structure and restraints due to function and three methods have been suggested by Chelliah (2004). The three methods are given as follows: 1. The first method requires identification of residues that have a higher degree of conservation than expected; 2. The second uses information theory to predict the overall amino acid substitution patterns and 3. The third identifies residues that have highly conserved positions with superposed three-dimensional structures of proteins (Chelliah et al, 2004). Although structures of proteins could be adequately identified and simulated with their gene sequences, protein functions could be described at different levels. Chelliah et al differentiate between molecular, cellular and physiological function of protein sequences. The molecular function of an enzyme depends on its specificity and the reaction which it catalyses; the cellular function may depend on its spatio- temporal expression in the cell, or how the functions are recognised within the spatial and temporal limits of a cell and the physiological function is characterised by the organ in which the expressing cells are found (Chelliah et al, 2004). The method used in Chelliah et al's (2004) study is clustering of residues in three-dimensional space and applying it to a set of well characterised protein families to identify the functional sites. The aim was to predict functional sites of intermolecular interaction and highlight the importance of both structural and functional restraints in protein sequences. In a related study, Gold and Jackson (2005) have recently emphasised on the importance of both protein structure and functional analysis. They suggest that the rapid growth of protein structural data and the emergence of the structural genomic projects have increased the need for automated structure analysis as well as show the need for new tools in prediction of functional relationships. Small molecule recognition has been considered as critical to the functions of protein so the determination of ligand binding site similarity is also important in understanding ligand interactions allowing functional classification and analysis. Gold and Jackson present a binding site database and when given a protein ligand binding site a rapid retrieval of other binding sites is also possible. These binding sites also showed similar structure independent of overall sequence or the fold similarity. Ligand binding sites are capable of indicating common binding modes and recognition of similar molecules allowing inference of function from protein or providing other evidence of common functions where fold similarity is known. The study highlights the need for detailed study of molecular recognition of structure based ligand designs to understand ligand cross reactivity. The paper shows the atomic similarity in structures between distant fold relatives and also between unrelated proteins (Gold and Jackson, 2005). As a result of identifying these similarities, unclassified proteins are assigned to structural superfamilies and sequence similarities were also identified. If sequence similarity failed to find significant matches, correct assignments were also possible showing the use of binding site comparisons for newly determined proteins. Thus in this study, both structural and functional aspects of protein analysis and ligand binding sites were used to identify underlying protein relationships. Heeb et al (2005) performed a functional analysis of post-transcriptional regulator RsmA to reveal a novel RNA-binding site. The RsmA family of RNA binding proteins are post transcriptional regulators that mediate extensive changes in gene expression in bacteria. This suggests the functional attributes of these binding proteins resulting in changes in gene expression. These proteins bind to and affect translation rate of target mRNAs which is another function modulated by regulatory RNAs. For new insights into the nature of this protein-RNA interaction, Heeb et al (2005) use x-ray crystallography to solve the structure. From the structure based alignments of RsmA protein family from diverse bacteria, the key amino acid residues involved in RNA binding were identified. The authors concluded that RsmA incorporates a novel class of RNA binding site and differs substantially from some RNA binding proteins. Heeb et al's study shows the importance of using a functional analysis to identify novel binding sites. The research objectives of this study showing the importance of studying the functional aspects of protein analysis gets further support from Heeb et al's study. The functional roles of amino acid residues in gene product or gp18 in contractile tail sheath protein of bacteriophage T4, were determined using the mutation sites and amino acid replacements of mutants and distinct phenotypes. This study was made by Takeda et al (2004) who used amber mutants for amino acid insertion by host amber suppressor cell strains. The study found that mutants giving rise to a particular phenotype could be mapped in a particular region along the polypeptide chain. Thus, all amino acid replacements in the cold-sensitive mutants and the heat-sensitive mutants were mapped in a limited region in the C-terminal domain. On the other hand, all the temperature-sensitive mutants and carbowax mutants were mapped in the N-terminal protease-resistant domain. The results suggested that the C-terminal region of gp18 is important for contraction and assembly as it contained all cold sensitive and heat sensitive mutants, whereas the N-terminal protease-resistant domain forms the protruding part of the tail sheath. The unique functional role of the amino acid replacements in the mutants also suggested a distinct phenotype to show close relations of phenotype and functionality (Takeda et al, 2004). This study relates directly to our foundational study on structural and functional evolutionary restraints by highlighting the importance of the basic relationship between phenotype and functionality. Considering simulation of protein structures, Zagrovic et al (2002) discuss a study in which the atomistic details of a fast folding 36 residue alpha helical protein from the villin headpiece were simulated. The total simulation time exceeded in several orders of magnitude compared with the previous simulations. The authors started from an extended state, obtained an ensemble of folded structures and emphasized that the folding mechanism of villin is consistent with the hydrophobic collapse view of folding. The 'hydrophobic collapse model' hypothesises that the protein conformation forms by rearrangement of a compact collapsed structure. Folding mechanisms however vary significantly according to protein size, stability and structure. The molecule collapsed non-specifically and reduced the size of conformational space. The authors state that the conformational search in the collapsed state appeared to be rate limited by the formation of the aromatic core. In a significant fraction of simulations, the C-terminal phenylalanine residue packed improperly with the rest of the hydrophobic core. The breaking of this interaction has been considered as rate determined in the course of the folding and the paper suggests that on the basis of the simulations, the folding rate of the villin was considered to be approximately 5s. The average features of the folded ensemble obtained by simulation were analysed to suggest that the mean folded structure is more similar to the native fold than an individual folded structure. Zagrovic and colleagues emphasise on the need for simulating ensembles of molecules and averaging the results in an experiment-like fashion in order to do a meaningful comparison between simulation and experimentation. The authors use their findings to compare with existing protein theories. The results of the findings have demonstrated that (1) the computational methodology exists to simulate the multi-microsecond regime using distributed computing, and (2) that potential sets used to describe inter-atomic interactions may be sufficiently accurate to reach the folded state, at least for small proteins (Zagrovic et al, 2002). The importance of inter-atomic interactions being fundamental in determining the folded state of protein, the implications of this study relates to the structural analysis of protein using interactional patterns. In an influential study, Levitt (1976) reports on the treatment of protein conformations. Here the concept of time-averaged forces was used to simplify conformational energy calculations on globular proteins. The description was given of a simplified molecular geometry, parameterization of suitable force fields, energy minimisation procedures and techniques for escaping local minima in which the protein's conformational space is achieved by converging larger regions into single points and the local minimum energy structure is found. Native conformations were done on pancreatic trypsin inhibitor to show that the simplifications represent the stable native conformation of this globular protein. Simulated folding of pancreatic trypsin inhibitor from open chain conformations were shown to give compact calculated conformations. Simulations on protein folding were done under a variety of conditions and the study discusses the actual in vitro folding process along with relevance of folding simulations (Levitt, 1976). The techniques seem to have many potential applications and include enzyme substrate binding, changes in protein tertiary structure and protein-protein interactions. Determination of protein-protein interactions, parameterization of force fields and locating the protein's conformational space seem to be important in determining both the structural and functional evolutionary constraints through simulation programs. Considering interaction sites as the basis of their study, Lin et al (1999) used the bacteriophage DNA packaging results from an ATP driven translocation. Mutant analysis and sequence localization were done to determine the functional domains of bacteriophage T4 terminase and portal gene product 20 within the structural genes. The interaction regions of portal vertex and large terminase subunits were determined by genetic, biochemical, and immunological studies. Bacteriophage with linear DNA genomes packages the DNA into pre-assembled pro-capsids through a vertex containing a ring of portal protein (Moore et al, 2002). This is known as the portal vertex and serves as a recognition site for packaging enzymes. The specificity of the interactions was tested with T4HOC which is a highly antigenic outer capsid protein and peptide sequences were displayed. The specified nature of portal protein terminase interaction sites were supported with co-immuno precipitation using purified terminase. The portal protein interaction site was localised to 28 residues in the central position of the linear sequence and two separate regions of the terminase were found to interact through hydrophobic contacts at the portal site specifically judged by inter genic portal terminase suppressor mutations. The terminase was found to interact with the portal peptide and since it was not accessible in its native structure an intimate association and interaction of gene products 20 and 17 have been suggested to show that such interactions internalized terminase regions within the portal complex (Lin et al, 1999). Thus extensive, association and interaction have been suggested in protein complexes to suggest mutations and localization changes highlighting the importance of intermolecular interactions as an indicator of functionality. Various functional activities have been studied using computational tools and simulation frameworks. For example Yang and Sharp (2004) show that the random network model of water quantitatively describes the different hydration heat capacities of polar and apolar solutes in terms of differential distortions of the water-water hydrogen bonding. The method of hydration analysis is applied here to study hydration of thermal hysteresis protein from eel pout and three mutations at residues. The study revealed the functional attributes of hydrogen bonding showing that the binding surface of active protein strongly enhances the water tetrahedral structure and promotes ice-like hydration. The study concluded that the specific shape, residue size and clustering of both polar and the apolar solutes are essential for the binding surface to recognize, and preferentially interact with nascent ice crystals forming in liquid water (Yang and Sharp 2004). Protein sequences thus have to be analysed both in terms of structural gene patterns and in terms of functional changes through intermolecular interactions. The focus of our study is thus based both on the structural and the functional aspects of protein analysis. The methods of simulation are based on Crescendo, Amber or CHARMM as bioinformatics programs to quantify unusual residue in the protein considering aspects of hydrogen bonding as primary features of the models used. Conclusion The literature review was based on a comprehensive study of the different force fields used as bioinformatics programs determining protein analysis mechanisms. The various programs discussed here are models of Amber, CHARMM, Crescendo and OPLS. The various protein analysis programs have been discussed in some depth along with the role of hydrogen atoms in these computational tools. The importance of simulation of atomic structures and proteins in terms of gene sequences were studied suggesting the three levels of bioinformatics at which all studies are based. From the first level of simple gene analysis, determining complete genome to obtaining functional data involving genes and genomes, the three major steps at which protein analyses are done have been delineated. Determining structural and functional evolutionary relationships between proteins, locating protein-protein interactions, predicting protein structure and functions and clustering proteins for efficient classification are some of the motivating factors for protein analysis. The various studies drawn upon suggest the importance of distinguishing the structural and functional evolutionary restraints on protein sequences. The functional aspects of protein analysis definitely points towards the dynamics of an inherent interactional pattern that have been elaborated using several other recent studies by Zagrovic et al (2002), Lin et al (1999) and others. The simulation techniques used in protein analysis have been exemplified with the help of studies on distributed computing and random network model of water using the water -water hydrogen bonding to show interactional patterns. Several other related simulation techniques and the role of protein-protein interactions have been discussed to present a comprehensive analysis of the application of computational and bioinformatics programs on protein analysis. In the next part of our analysis, we will deal with the methodology for this study and will subsequently set out the research objectives. Aims and Objectives The project is aimed at analysing the structural and functional evolutionary restraints of protein families using bioinformatics simulation programs such as Crescendo and CHARMM to perform this analysis. Thus protein analysis involves three important phases 1. Identification of structural and functional components 2. Categorisation of protein families according to these components 3. Determining evolutionary relationships of protein structural and functional characteristics. Considering these general aims of protein analysis, this study is based on identification and grouping of both the structural and the functional evolutionary restraints in proteins using computational techniques and bioinformatics programs to aid the purposes of the study. The main objective is to highlight the importance of functional data identified by interactional patterns (in Chelliah et al, 2004) and we suggest that despite the importance given to the structural analysis of proteins in current biochemical research studies, the role of functional analysis may be fundamental to the study of proteins. Considering this objective which tilts towards focusing on the importance of a 'functional' role of evolutionary restraints and identifies the importance of functional data and analysis and intermolecular interactions, our literature review has been geared to highlight research studies on functional attributes of protein analysis within the special context of simulation tools and bioinformatics programs. A thorough evaluation of available bioinformatic programs has been done showing the attributes, strengths and weaknesses of each such program. We use CHARMM and Crescendo programs in this study to test our hypothesis on the central role of functional evolutionary restraints identified through intermolecular interactional patterns in protein analysis. REFERENCES Amber Group (Accessed 2005) - The Amber Molecular Dynamics Package. http://amber.scripps.edu/. Arqus, Didier G. Fallot Jean-Paul and Michel Christian J. (1996) Identification of several types of periodicities in the collagens and their simulation. International Journal of Biological Macromolecules, Volume 19, Issue 2, August, Pages 131-138. Bioinformatics Web (Accessed 2005) - BIW Online Resource. http://www.geocities.com/bioinformaticsweb/. Chelliah, Vijayalakshmi Chen, Lan Blundell Tom L. and Lovell Simon C. (2004) Distinguishing Structural and Functional Restraints in Evolution in Order to Identify Interaction Sites. Journal of Molecular Biology, Volume 342, Issue 5, 1 October, Pages 1487-1504. Chemistry at HARvard Macromolecular Mechanics (Accessed 2005) - CHARMM. http://www.charmm.org/. Crystallography and Bioinformatics, Deptartment of Biochemistry, University of Cambridge (Accessed 2005) - Web Servers. http://www-cryst.bioc.cam.ac.uk/servers.html Dandekar Thomas and Argos Patrick (1996) Identifying the Tertiary Fold of Small Proteins with Different Topologies from Sequence and Secondary Structure using the Genetic Algorithm and Extended Criteria Specific for Strand Regions. Journal of Molecular Biology, Volume 256, Issue 3, 1 March , Pages 645-660. Desrosiers Daniel C. and Peng Zheng-yu (2005) A Binding Free Energy Hot Spot in the Ankyrin Repeat Protein GABP Mediated Protein-Protein Interaction. Journal of Molecular Biology, Volume 354, Issue 2, 25 November 2005, Pages 375-384. Gaeta, Bruno A.(Bruno Andre) (1997) ANGIS Bioinformatics handbook.Vol.2,Basic Bioinformatics Techniques. Collingwood, VIC :CSIRO Australia,c1997. Gayathri, P. Satheshkumar, P. Prasad, S. K. Nair, Smita Savithri H.S. and Murthy M.R.N. (2005) Crystal structure of the serine protease domain of Sesbania mosaic virus polyprotein and mutational analysis of residues forming the S1-binding pocket. Virology, 13 December. Gold Nicola D. and Richard M. Jackson (2005) Fold Independent Structural Comparisons of Protein-Ligand Binding Sites for Exploring Functional Relationships. Journal of Molecular Biology, 1 December 2005. Heeb, Stephan Kuehne, Sarah A. Bycroft, Mark Crivii, Sorana Allen, Mark D. Haas Dieter Cmara, Miguel and Williams Paul (2005)Functional Analysis of the Post-transcriptional Regulator RsmA Reveals a Novel RNA-binding Site. Journal of Molecular Biology, 1 December 2005. Hizlan, Dilem Mishima, Masanori Tittmann, Gross, Peter Heinz Glotzer Michael and Hoenger Andreas (2005) Structural analysis of the ZEN-4/CeMKLP1 motor domain and its interaction with microtubules. Journal of Structural Biology, 29 November. Ito, Genta Ariga, Hiroyoshi Nakagawa Yasuhito and Iwatsubo Takeshi (2005/2006) Roles of distinct cysteine residues in S-nitrosylation and dimerization of DJ-1. Biochemical and Biophysical Research Communications, Volume 339, Issue 2, 13 January 2006, Pages 667-672. Khalili, Mey Liwo Adam and Scheraga Harold A. (2005/2006) Kinetic Studies of Folding of the B-domain of Staphylococcal Protein A with Molecular Dynamics and a United-residue (UNRES) Model of Polypeptide Chains. Journal of Molecular Biology, Volume 355, Issue 3, 20 January 2006, Pages 536-547. Lee, Jun Hyuck Sung-Tae Yang, Seong-Hwan Rho, Young Jun Im, Soo Young Kim, Young Ran Kim, Mun-Kyoung Kim, Gil Bu Kang, Jae Il Kim, Joon Haeng Rhee and Soo Hyun Eom (2005/2006) Crystal Structure and Functional Studies Reveal that PAS Factor from Vibrio vulnificus is a Novel Member of the Saposin-fold Family. Journal of Molecular Biology, Volume 355, Issue 3, January 2006, Pages 491-500. Lee Susan D. and Alani Eric (2005/2006) Analysis of Interactions Between Mismatch Repair Initiation Factors and the Replication Processivity Factor PCNA. Journal of Molecular Biology, Volume 355, Issue 2, 13 January 2006, Pages 175-184. Lengauer, Thomas. Bioinformatics :from genomes to drugs. Vol. 1, Basic Technologies. Methods and principles in medicinal chemistry. vol 14 Weinheim: Wiley-VCH,c2002. Lesk, Arthur M. (2003) Introduction to Bioinformatics. Oxford University Press, Oxford. 283pp. Levitt Michael (1976) A simplified representation of protein conformations for rapid simulation of protein folding. Journal of Molecular Biology, Volume 104, Issue 1, 14 June Pages 59-107. Lin, Hsingchi Venigalla B. Rao and Lindsay W. Black (1999) Analysis of capsid portal protein and terminase functional domains: interaction sites required for DNA packaging in bacteriophage T4. Journal of Molecular Biology, Volume 289, Issue 2, 4 June Pages 249-260. Moore SD, Prevelige PE Jr. (2002) Bacteriophage p22 portal vertex formation in vivo. J Mol Biol. Feb 1;315(5):975-94. Novotn, Ji Robert Bruccoleri and Martin Karplus (1984) An analysis of incorrectly folded protein models : Implications for structure predictions. Journal of Molecular Biology, Volume 177, Issue 4, 25 August Pages 787-818. Olson Wilma K (1996) Simulating DNA at low resolution. Current Opinion in Structural Biology, Volume 6, Issue 2, April, Pages 242-256. Oostenbrink C, Soares TA, van der Vegt NF, van Gunsteren WF. (2005) Validation of the 53A6 GROMOS force field. Eur Biophys J. Jun;34(4):273-84. Epub 2005 Apr 1. Ottosson, Lars Janosch Hennig, Alexander Espinosa, Susanna Brauner, Marie Wahren-Herlenius and Maria Sunnerhagen (2005/2006) Structural, functional and immunologic characterization of folded subdomains in the Ro52 protein targeted in Sjgren's syndrome. Molecular Immunology, Volume 43, Issue 6, February, Pages 588-598 Oxford Journals (Accessed 2005) - Bioinformatics Database/Journals. http://bioinformatics.oxfordjournals.org/ Ponder JW, Case DA. (2003) Force fields for protein simulations. Adv Protein Chem.;66:27-85. Singh, U. T. Sun, T. Larsson, R.W. Elliott, G. Kostka and R.H. Fundele (2005)Expression and Functional Analysis of Fibulin-1 (Fbln1) During Normal and Abnormal Placental Development of the Mouse. Placenta, 9 December. Takeda, Shigeki Makoto Suzuki, Takahito Yamada, Manabu Kageyama and Fumio Arisaka (2004) Mapping of functional sites on the primary structure of the contractile tail sheath protein of bacteriophage T4 by mutation analysis. Biochimica et Biophysica Acta (BBA) - Proteins & Proteomics, Volume 1699, Issues 1-2, 1 June, Pages 163-171. Tsai, Chung-Jung Lin, Shuo Liang Wolfson Haim J. and Nussinov Ruth A (1996) Dataset of Protein-Protein Interfaces Generated with a Sequence-order-independent Comparison Technique. Journal of Molecular Biology, Volume 260, Issue 4, 26 July 1996, Pages 604-620. Winckler, Thomas Christine Trautwein, Christina Tschepke, Christin Neuhuser, Ilse Zndorf, Peter Beck, Gnter Vogel and Theodor Dingermann (2001) Gene function analysis by amber stop codon suppression: CMBF is a nuclear protein that supports growth and development of Dictyostelium amoebae. Journal of Molecular Biology, Volume 305, Issue 4, 26 January, Pages 703-714. Yang Cheng and Sharp Kim A (2004) The mechanism of the type III antifreeze protein action: a computational study. Biophysical Chemistry, Volume 109, Issue 1, 1 April, Pages 137-148. Zagrovic Bojan, Snow, Christopher D. Shirts Michael R. and Pande Vijay S.(2002) Simulation of Folding of a Small Alpha-helical Protein in Atomistic Detail using Worldwide-distributed Computing. Journal of Molecular Biology, Volume 323, Issue 5, 8 November, Pages 927-937. Read More