Current Bioinformatics - Volume 7, Issue 4, 2012

About 15 years ago, the steered molecular dynamics (SMD) was used to probe binding of ligand to biomolecule surfaces but in terms of drug design this approach has only recently attached attention of researchers. The main idea of using SMD to screen out leads is based on the hypothesis that the larger is the force needed to unbind a ligand from a receptor the higher its binding affinity. Thus, instead of binding free energy, the rupture force defined as the maximum on the force-time/displacement profile, is used as a score function. In this mini-review, we discuss basic concepts behind the experimental technique atomic force microscopy as well as SMD. Experimental and theoretical works on the application of SMD to the drug design problem are covered. Accumulated evidences show that SMD is as accurate as the molecular mechanics-Poisson-Boltzmann surface area method in predicting ligand binding affinity but the former is computationally much more efficient. The high correlation level between theoretically determined rupture forces and experimental data on binding energies implies that SMD is a promising tool for drug design. Our special attention is drawn to recent studies on inhibitors of influenza viruses.

In order to obtain structural information about intermolecular interactions between a protein target and a drug we could either solve the structure by experimental techniques (protein crystallography or nuclear magnetic resonance), or simulate the protein-drug complex computationally. Molecular docking is a computer simulation methodology that can predict the conformation of a protein-drug complex, with relatively high accuracy when compared with experimental structures. Although a plethora of algorithms has been applied to the problem of molecular docking simulation, recent results show that the most successful approaches are those based on evolutionary algorithms. Evolution as a source of inspiration has been shown to have a great positive impact on the progress of new computational methodologies. In this scenario, analyses of the interactions between a protein target and a drug can be simulated by these evolutionary algorithms. These algorithms mimic evolution to create new paradigms for computation. This review provides a description of evolutionary algorithms and applications to molecular docking simulation. Special attention is dedicated to differential evolutionary algorithm and its implementation in the program molegro virtual docker. Recent applications of these methodologies to protein targets such as acetylcholinesterese, cyclin-dependent kinase 2, purine nucleoside phosphorylase, and shikimate kinase are described.

Amantadine is a specific anti-influenza a drug that inhibits viral replication by binding to the M2 channel and preventing proton conductance. The increasing resistance to amantadine in strains of the influenza A virus that infect both animals and humans has been highlighted frequently. Resistance is usually caused by one of several single mutations in the M2 channel, but variants with double mutations have also been reported. Attempts to develop alternative inhibitors of the M2 channel that are effective against the resistant mutants have been unsuccessful, mainly because of the lack of information on the precise mode of inhibitor binding. This review summarizes the advances made in determining the mechanisms of action of amantadine and the development of novel inhibitors of the M2 channel during the past 2 years.

The shikimate pathway is found in microorganisms, fungi, plants and also in several apicomplexan parasites. This metabolic pathway consists of seven enzymes and converts the primary metabolites phosphoenolpyruvate and erythrose-4-phosphate to chorismate, the last common precursor for the three aromatic amino acids Phe, Tyr, and Trp and other aromatic compounds. The significance of targeting the enzymes of this pathway as selective targets for anti microbial drug design involves the fact that they are essential for microbes but absent in humans. In present scenario, the emergence of multi-drug resistance in pathogenic bacteria and herbicide resistance in weeds is of great clinical and agro-economical concern. Therefore in this review, we did the comparative sequence and threedimensional structure analysis of these enzymes from various microorganisms and plants for structure-function analysis, motif search, common structural signatures of active site and elucidation of regulation mechanisms. Also, the available structures of five shikimate pathway enzymes from M. tuberculosis, a dreadful microorganism, which causes 1.5 million deaths per year, have been comparatively analyzed with other reported homologous structures. To get the structural insight of remaining two shikimate pathway enzymes (dehydroquinate synthase and shikimate-5-dehydrogenase) of M. tuberculosis we did molecular modeling to find out key active site residues. These studies can further be proven helpful in designing novel structure based antimicrobial drugs.

Caspases belong to the family of cysteinyl aspartate–a specific proteases which control the programmed cell death process, or apoptosis. In this paper, we have performed a structural bioinformatics analysis of the conserved domains and residues, WebLogo generation and active sites identification related to apoptosis activator and apoptosis executioner caspase-cascades. Here, we have also shown conservation patterns of backbone structures of activator and executioner caspase-cascades. It has been noted that the numbers of highly conserved amino acid residues are very high in caspase-12 (36 aa) and low in caspase-7 (18 aa). We have observed that highly conserved amino acids residues like LYS154, PRO155, LYS156 are present in caspase-3 and caspase-6. In apoptosis and executioner caspases, these amino acids may play an active role. From WebLogo, it has been observed that the stack height is very low between the sequences 231 to 240; 2.3 bits stack height has been observed in 1st sequence position and 236th position where WebLogo stack height is very low. We have identified 10 active sites in caspase-3, caspase-6, caspase-7 which may be helpful in drug development using caspase-cascades. Here, we have also performed literature survey about the drug development using caspase-cascades.

Caspases are very important molecules which are playing a key role for apoptosis. Deregulation of apoptosis contributes to the pathogenesis of many human diseases. Therefore, the regulation of this protein can be controlled for the therapeutic purpose. To determine molecular evolution, in silico of proteins is popular a method for certain laboratories and biotechnology companies. It may help to detect that mutations often occur far away from the active site of the protein. It can give us an insight thinking about drug development. Using sequence analysis and phylogenetic approach, we have described about different human caspases and about their origin in terms of ancestral relationship. It was envisaged using the tools of bioinformatics. Among the fourteen mammalian caspases defined, we are able to make use of twelve human caspases, whose data is publicly available. It is evident from the data studied that human caspases 4 and 5 share the same origin in comparison to human caspase 1 and caspase 12, irrespective of the fact that both share quite high level of similarity. Although, human caspase related ancestral aspect had been studied earlier but the variation which seems to be quite peculiar in this study is that the executioner, caspase 3 shares a remarkable high level of similarity with caspase 7 but this is not applicable to human caspase 6, the other member of executioner group. Human caspases 3 and 7 were seen to have similar substrate specificity but it was not evident in terms of the origin. Our findings are assumed to play a significant role in the studies of programmed cell death, inflammatory responses and for scholarly studies in the near future.

Orphan nuclear receptors (oNRs) provide huge opportunities for the discovery of new drug targets. Identifying novel and perhaps unexpected types of ligands for oNRs may gain insight into potentially new principles of physiology. Recently, the network-based approaches are playing an increasingly important role in identifying and validating novel ligands for nuclear receptors. This review describes current progresses in network-based approaches for ligand prediction and discusses their strengths and some of the underlying difficulties.

Sequence classification is one of the most fundamental machine learning tasks in computational biology nowadays. With the wide availability of large corpora of annotated sequences, the use of supervised learning techniques can greatly speed up the process of identifying new sequences sharing certain function or properties. Many methods have been proposed over the years and we hope to provide an introduction to some of the more prominent ones by focussing on protease cleavage prediction: a typical representative of this class of problem. The variety of proteolytic action modes between cysteine-proteases covers a broad range of complexity level and feature specificity, illustrating the strengths and limitations of the different machine learning techniques used on them. This review briefly introduces the particulars of predicting cleavage by calpains and caspases. We then offer some general practical considerations on treating sequences for use with machine learning algorithms, before covering specific methods. The methods presented range from basic position-based statistical models to more technically advanced methods such as Markov models or kernel-based algorithms, as well as methods with more restricted goals such as decision trees. With each family of algorithms, examples of implementations are introduced and their performances compared, along with particular strengths and weaknesses. With this review, we aim to provide useful elements of decision toward choosing an existing method or developing a new one, based on the complexity and specific needs of a given sequence classification problem.

If structures of proteins of whole organisms were available, metabolomic models could be developed, drug targets could be identified, issues of affinity versus specificity could be sorted out and side effects and toxicity brought under control etc. all with greater levels of reliability. Advances in whole genome sequencing projects, annotation algorithms, growing protein sequence information with over half a million entries in the UniProtKB/Swiss-Prot database, progresses in structure based lead molecule design methodologies do uphold this optimism. However, x-ray and NMR structures of less than 15% of the protein sequences are available in RCSB protein data bank. The diverging gap between sequence and structure calls for immediate in silico solutions. The biennial community wide structure prediction (CASP) experiments have considerably catalyzed structure prediction attempts world-wide and accuracies of computational models are continually increasing. While ab initio models have crossed the 100 amino acid limit, it is still some way from the average sized human protein (~ 350 residues). Homology models which rely on the RCSB structures and the axiom that similar sequences adopt similar structures have been extremely powerful in providing high resolution structures limited only by sequence similarities. With dwindling similarities of query sequences with knowledge bases, newer ab intio / homology hybrid approaches are being explored to bring the structure prediction problem within the realm of feasibility in near future particularly for soluble proteins. The case of membrane bound proteins is still refractory. This review takes a stock of current protein tertiary structure prediction algorithms highlighting the problem areas to overcome and promises thereof.

Designing ontology to represent gene function is of vital importance for meeting the major challenge of integrating sequence data with the increasing amount of data from functional analyses of genes. Given that genes are expressed in temporally and spatially characteristic patterns, their products quite often reside in specific cellular compartments and may be part of one or more multi-component complexes. Genes may have more than one product and the products are functionally distinct. An overall strategy elucidating how an ontology-based gene function may be implemented using genomic databases is herein dissected. Knowing that gene products possess one or more biochemical, physiological or structural functions, the present strategy is suggested to lead towards physiological models. A review of the features of the currently available software tools for the implementation of the considered strategy is presented.

Transcriptional regulation is a key step to control the level of mRNA formed. Recent view of transcriptional regulation has evolved from a one-dimensional mode, i.e. RNA Polymerase II assembles with general transcription factors, and cis-regulatory elements (CREs) interact with transcription factors, to a much complex multiple-dimensional mode, involving combinatorial interactions between transcription factors and regulatory sequences, chromatin structure, histone modifications, DNA methylation. High throughput experimental technologies, such as array-based ChIP-chip and sequencing-based ChIP-seq, have been developed to survey in vivo transcription factor binding sites and histone modifications. Despite many efforts have been made to analyze and interpret the data, challenges remain in many aspects of both experimental protocols and computational analyses. For example, how to determine the optimized number of PCR cycles? How to normalize multiple datasets from multiple experiments? How to utilize the large number of unmapped and multiple mapped tags in ChIP-seq experiment? This review focuses on issues emerged in high throughput data processing and discusses advantages and disadvantages of various strategies.

Emerging and re-emerging viral infections are a threat to human health and a cause of global concern. Several viral vaccines have been successfully developed using conventional methods. However, there are many viruses for which vaccines need to be developed on priority basis. Furthermore, the challenges viz, varying efficacy of existing vaccines also need to be addressed as viruses are known to evolve at a higher rate as compared to other species. Under this scenario, availability of whole genome sequences of pathogens has brought a paradigm shift and reversed the process of vaccine development, which is termed as reverse vaccinology approach. The advents of next generation sequencing technologies coupled with pan-genomic approaches, offer unprecedented opportunities for data driven, knowledge-based approaches for rational design of viral vaccines. Reverse vaccinology approach begins with analysis of genomic data and culminates into identification and prioritization of a few tractable vaccine candidates. The genomic sequence of a viral pathogen is processed through various sequence and structure-based analyses to identify an ensemble of epitopes, which not only reduces time required for discovery but also helps in eliminating quite a few in vitro screening steps. The field leverages on tools & techniques of bioinformatics, immunoinformatics and comparative genomics, which are not only independent domains of research, but also offer a distinct advantage in designing vaccines. A detailed account of the state of art resources and methods in each of these areas are reviewed and presented to substantiate and highlight their roles in designing viral vaccines in the post genomics era.

Bunyaviridae family consists of vector borne lethal viruses, stands out as the largest virus family with its 350 members. One such virus of this family, Crimean- Congo hemorrhagic fever virus (CCHFV) is transmitted through bites of ixodid ticks or by direct contact with blood from infected animals. Crimean-Congo haemorrhagic fever (CCHF) is a severe disease in humans which is endemic in large parts of the world with a high mortality rate. This virus could also be used as a bioterrorism agent due to its human-to-human transmission with no specific therapy. The pathogenicity factor of CCHFV is unexplored due to the lack of animal models. CCHFV, being an RNA virus, is able to mutate rapidly hence preventing the development of effective therapy against it. Till now ribavarin is the only available drug for supportive treatment but has many side-effects. New technologies like RNA interference have emerged as a solution for epidemics of CCHF. RNAi is a sequence specific approach, has been used successfully against different pathogens. This review focuses on designing and application of RNAi with emphasis on the role of bioinformatics for the anti CCHFV therapeutic development strategy.

Malaria is a serious infectious disease that causes over one million deaths yearly. It is caused by a group of protozoan parasites in the genus Plasmodium. No effective vaccine is currently available and the elevated levels of resistance to drugs in use underscore the pressing need for novel antimalarial targets. In this review, we survey omics centered developments in Plasmodium biology, which have set the stage for a quantum leap in our understanding of the fundamental processes of the parasite life cycle and mechanisms of drug resistance and immune evasion.

Since most molecular studies on death of cells in tissues have been carried out on isolated cell populations due to known difficulties manifested by interactions with surrounding cells, a novel means of investigating general principles governing cellular functions under oxidative stress conditions is needed in order to shed more light on the background of cancer disease. It is believed that relevant signal transmission may be discovered by transition from molecular to modular cell biology. Systems-level kinetic models are thus expected to explain dynamic behavior and go far beyond the static pictures of the topologies of the signaling pathways. The outline of this review is to feature several representative problems, based on combined - experimental and systems biology studies over the last few years, with a particular emphasis both on the elucidation of how cells interpret the same signal stimulation in distinct fashions (cell death vs. cell survival) and on the identification of signaling molecules with therapeutic relevancy. The origin of oscillations in such molecular mechanisms under oxidative stress conditions and implications of these oscillatory non-linearities for the development of successful therapies are discussed.

Sequence alignment is a basic field in bioinformatics, especially the sequence alignment of remotely homologue proteins is a hot spot. In our previous work, we developed a new score matrix named transformation matrix which can greatly enhance the quality of the alignment of distant protein sequences. Here, by using the transformation score matrix, we assessed the statistical significance of the local sequence alignment. Compared with the traditional score matrix, the local sequence alignment method has the following features: (i) The optimal alignment scores approximately follow a normal distribution. (ii) The distribution is closely related with N, which represents the length of two sequence alignments but not the lengths of the two sequences being compared. Therefore, for a pair of two aligned protein sequences, we can calculate the P-value based on the N and the optimal alignment score.

MicroRNAs (miRNAs) are short RNA molecules that regulate the post-transcriptional expression of their target genes – messenger RNAs (mRNAs). Although miRNAs are known to down-regulate the translation or the cleavage of mRNAs, the miRNA-mediated regulation mechanisms still remain largely unknown. Therefore, in order to achieve a better understanding of the biological function of the miRNAs, a critical step is to identify miRNA targets. At present, since the experimental identification of miRNA targets is still difficult, investigators rely mainly on computational methods to identify miRNA targets, and numerous computational methods are currently available. The computational identification of miRNA targets has emerged as a significant field in miRNA target research. In this paper, we summarize the knowledge about the computational identification of miRNA targets, including the principles of miRNA targeting, computational methods for miRNA target prediction, and the databases required for miRNA targeting. Furthermore, we make a comparison analysis of several representative computational target prediction methods in this work.