News Article | April 19, 2017
More effective cancer treatments are likely to emerge from the drug development pipeline. Cancer drug discovery hinges on identifying and characterizing binding pockets in target proteins. Typically, this evaluation uses computational techniques that rely on static protein structures. However, proteins have an inherent flexibility that causes a tendency to change shape upon contact with the drugs. Certain binding pockets remain undetectable unless they interact with an appropriate substance and, therefore, are missed by conventional simulations. These hidden pockets, however, are usually water-repelling or hydrophobic sites that only open when there are low polarity substances. To tackle this, Yaw Sing Tan and Chandra Verma from the Bioinformatics Institute have developed a probe-based method called ligand-mapping molecular dynamics (LMMD). They used this technique to seek hidden binding pockets in the anticancer target protein MDM2. The resulting predictions were experimentally validated by long-standing collaborators from A*STAR's p53 Laboratory and Institute of Chemical and Engineering Sciences as well as structural biologists from Newcastle University, UK. Tan explains that initially he had designed this probe-based method for another target protein and successfully used it to find a hidden binding pocket that stayed closed in conventional simulations. "We then decided to apply this approach to MDM2 to see if we could discover any previously unknown binding sites that could enhance the potency of existing MDM2 inhibitors," he adds. Using benzene molecules as hydrophobic pocket detection probes, the researchers computationally identified two new binding sites on MDM2. "We were excited to see that these sites lie very close to the binding pocket of the tumor suppressor protein p53," says Tan. Furthermore, the researchers expect the newly found sites to lead to more potent stapled peptides—these are amino acid helices chemically stabilized by a hydrocarbon chain that have recently emerged as powerful p53 activators. Consequently, they created stapled peptides from analogs known to tightly bind MDM2 and reactivate p53, and determined the affinity of these peptides to MDM2. Their simulations showed that the peptides bound MDM2 more strongly than p53 in the pockets and matched biophysical and X-ray crystallography experiments. "This method could be used to interrogate other anticancer protein targets to uncover novel binding sites that could be targeted for inhibition," says Tan. The team is now working to expand the reach of LMMD probes to other ligand types. Explore further: Novel use of fluorescent probe may lead to future anti-cancer drugs More information: Yaw Sing Tan et al. Benzene Probes in Molecular Dynamics Simulations Reveal Novel Binding Sites for Ligand Design, The Journal of Physical Chemistry Letters (2016). DOI: 10.1021/acs.jpclett.6b01525
Vaser R.,University of Zagreb |
Sovic I.,Ruder Boskovic Institute |
Nagarajan N.,Genome Institute of Singapore |
Sikic M.,University of Zagreb |
Sikic M.,Bioinformatics Institute
Genome Research | Year: 2017
The assembly of long reads from Pacific Biosciences and Oxford Nanopore Technologies typically requires resource-intensive error-correction and consensus-generation steps to obtain high-quality assemblies. We show that the error-correction step can be omitted and that high-quality consensus sequences can be generated efficiently with a SIMD-accelerated, partialorder alignment-based, stand-alone consensus module called Racon. Based on tests with PacBio and Oxford Nanopore data sets, we show that Racon coupled with miniasm enables consensus genomes with similar or better quality than state-of-theart methods while being an order of magnitude faster. © 2017 Vaser et al.
Shi Q.,University of Adelaide |
Cheng L.,Bioinformatics Institute |
Wang L.,Nanjing Forestry University |
International Journal of Computer Vision | Year: 2011
A challenging problem in human action understanding is to jointly segment and recognize human actions from an unseen video sequence, where one person performs a sequence of continuous actions. In this paper, we propose a discriminative semi-Markov model approach, and define a set of features over boundary frames, segments, as well as neighboring segments. This enable us to conveniently capture a combination of local and global features that best represent each specific action type. To efficiently solve the inference problem of simultaneous segmentation and recognition, a Viterbi-like dynamic programming algorithm is utilized, which in practice is able to process 20 frames per second. Moreover, the model is discriminatively learned from large margin principle, and is formulated as an optimization problem with exponentially many constraints. To solve it efficiently, we present two different optimization algorithms, namely cutting plane method and bundle method, and demonstrate that each can be alternatively deployed in a "plug and play" fashion. From its theoretical aspect, we also analyze the generalization error of the proposed approach and provide a PAC-Bayes bound. The proposed approach is evaluated on a variety of datasets, and is shown to perform competitively to the state-of-the-art methods. For example, on KTH dataset, it achieves 95.0% recognition accuracy, where the best known result on this dataset is 93.4% (Reddy and Shah in ICCV, 2009). © 2010 Springer Science+Business Media, LLC.
Koh Y.W.,Bioinformatics Institute |
Westerman K.,Tufts University |
Manzhos S.,National University of Singapore
Carbon | Year: 2015
We present an ab initio study of UF6 adsorption and vibrations on graphene derivatives: pristine single and double layer graphene, as well as (single layer) hydrogenated and fluorinated graphene. As the substrate results in a range of bonding strengths, from chemisorption to physisorption, both GGA density functional theory (DFT) and dispersioncorrected DFT (DFT-D) are used. The lowest adsorption energy, Eads, is of the order of 1.3/1.6 eV on single layer graphene, 1.2/1.5 eV on double layer graphene, 1.1/1.4 eV on graphane, and 0.1/0.3 eV on fluorographene, with DFT/DFT-D, showing that Eads can be tuned in a wide range by the choice of the substrate. The isotopic splitting in the vibrational spectrum of UF6 observed in vacuum is largely preserved in the adsorbed molecules. The existence of several adsorption configurations with competing Eads leads to overlaps in the vibrational spectra of isotopomers, but isotopomer-unique modes exist on all four surfaces. It may therefore be possible to cause desorption of a selected isotopomer by laser radiation, leading to isotopic separation between the surface and the gas. © 2014 Elsevier Ltd. All rights reserved.
Law Y.-N.,Bioinformatics Institute |
Wang H.,Microsoft |
Zaniolo C.,University of California at Los Angeles
ACM Transactions on Database Systems | Year: 2011
Most data stream management systems are based on extensions of the relational data model and query languages, but rigorous analyses of the problems and limitations of this approach, and how to overcome them, are still wanting. In this article, we elucidate the interaction between stream-oriented extensions of the relational model and continuous query language constructs, and show that the resulting expressive power problems are even more serious for data streams than for databases. In particular, we study the loss of expressive power caused by the loss of blocking query operators, and characterize nonblocking queries as monotonic functions on the database. Thus we introduce the notion of N B-completeness to assure that a query language is as suitable for continuous queries as it is for traditional database queries. We show that neither RA nor SQL are N B-complete on unordered sets of tuples, and the problem is even more serious when the datamodel is extended to support order-a sine-qua-non in data stream applications. The new limitations of SQL, compounded with well-known problems in applications such as sequence queries and data mining, motivate our proposal of extending the language with user-defined aggregates (UDAs). These can be natively coded in SQL, according to simple syntactic rules that set nonblocking aggregates apart from blocking ones. We first prove that SQL with UDAs is Turing complete. We then prove that SQL with monotonic UDAs and union operators can express all monotonic set functions computable by a Turing machine (N B-completeness) and finally extend this result to queries on sequences ordered by their timestamps. The proposed approach supports data streammodels that are more sophisticated than append-only relations, along with datamining queries, and other complex applications. © 2011 ACM.
Verma C.,Bioinformatics Institute
PLoS ONE | Year: 2015
The putative Major Facilitator Superfamily (MFS) transporter, SV2A, is the target for levetiracetam (LEV), which is a successful antiepileptic drug. Furthermore, SV2A knock out mice display a severe seizure phenotype and die after a few weeks. Despite this, the mode of action of LEV is not known at the molecular level. It would be extremely desirable to understand this more fully in order to aid the design of improved antiepileptic compounds. Since there is no structure for SV2A, homology modelling can provide insight into the ligandbinding site. However, it is not a trivial process to build such models, since SV2A has low sequence identity to those MFS transporters whose structures are known. A further level of complexity is added by the fact that it is not known which conformational state of the receptor LEV binds to, as multiple conformational states have been inferred by tomography and ligand binding assays or indeed, if binding is exclusive to a single state. Here, we explore models of both the inward and outward facing conformational states of SV2A (according to the alternating access mechanism for MFS transporters). We use a sequence conservation analysis to help guide the homology modelling process and generate the models, which we assess further with Molecular Dynamics (MD). By comparing the MD results in conjunction with docking and simulation of a LEVanalogue used in radioligand binding assays, we were able to suggest further residues that line the binding pocket. These were confirmed experimentally. In particular, mutation of D670 leads to a complete loss of binding. The results shed light on the way LEV analogues may interact with SV2A and may help with the ongoing design of improved antiepileptic compounds. © 2015 Lee et al.
Wongsurawat T.,Bioinformatics Institute |
Wongsurawat T.,Nanyang Technological University |
Jenjaroenpun P.,Bioinformatics Institute |
Kwoh C.K.,Nanyang Technological University |
And 2 more authors.
Nucleic Acids Research | Year: 2012
R-loop is the structure co-transcriptionally formed between nascent RNA transcript and DNA template, leaving the non-transcribed DNA strand unpaired. This structure can be involved in the hyper-mutation and dsDNA breaks in mammalian immunoglobulin (Ig) genes, oncogenes and neurodegenerative disease related genes. R-loops have not been studied at the genome scale yet. To identify the R-loops, we developed a computational algorithm and mapped R-loop forming sequences (RLFS) onto 66803 sequences defined by UCSC as 'known' genes. We found that ∼59 of these transcribed sequences contain at least one RLFS. We created R-loopDB (http://rloop.bii.a-star.edu.sg/), the database that collects all RLFS identified within over half of the human genes and links to the UCSC Genome Browser for information integration and visualisation across a variety of bioinformatics sources. We found that many oncogenes and tumour suppressors (e.g. Tp53, BRCA1, BRCA2, Kras and Ptprd) and neurodegenerative diseases related genes (e.g. ATM, Park2, Ptprd and GLDC) could be prone to significant R-loop formation. Our findings suggest that R-loops provide a novel level of RNA-DNA interactome complexity, playing key roles in gene expression controls, mutagenesis, recombination process, chromosomal rearrangement, alternative splicing, DNA-editing and epigenetic modifications. RLFSs could be used as a novel source of prospective therapeutic targets. © The Author(s) 2011. Published by Oxford University Press.
Masters B.C.,Bioinformatics Institute |
Fan V.,Bioinformatics Institute |
Ross H.A.,Bioinformatics Institute |
Ross H.A.,University of Auckland
Molecular Ecology Resources | Year: 2011
Species Delimitation is a plugin to the Geneious software to support the exploration of species boundaries in a gene tree. The user assigns taxa to putative species and the plugin computes statistics relating to the probability of the observed monophyly or exclusivity having occurred by chance in a coalescent process. It also assesses the within and between species genetic distances to infer the probability with which members of a putative species might be identified successfully with tree-based methods. © 2010 Blackwell Publishing Ltd.
Cheng L.,Bioinformatics Institute |
Gong M.,Memorial University of Newfoundland |
Schuurmans D.,University of Alberta |
Caelli T.,University of Melbourne
IEEE Transactions on Image Processing | Year: 2011
The authors examine the problem of segmenting foreground objects in live video when background scene textures change over time. In particular, we formulate background subtraction as minimizing a penalized instantaneous risk functional - yielding a local online discriminative algorithm that can quickly adapt to temporal changes. We analyze the algorithm's convergence, discuss its robustness to nonstationarity, and provide an efficient nonlinear extension via sparse kernels. To accommodate interactions among neighboring pixels, a global algorithm is then derived that explicitly distinguishes objects versus background using maximum a posteriori inference in a Markov random field (implemented via graph-cuts). By exploiting the parallel nature of the proposed algorithms, we develop an implementation that can run efficiently on the highly parallel graphics processing unit (GPU). Empirical studies on a wide variety of datasets demonstrate that the proposed approach achieves quality that is comparable to state-of-the-art offline methods, while still being suitable for real-time video analysis (≥75 fps on a mid-range GPU). © 2006 IEEE.
Batagov A.O.,Bioinformatics Institute |
Kurochkin I.V.,Bioinformatics Institute
Biology Direct | Year: 2013
Reviewers: This article was reviewed by Neil Smalheiser and Sandor Pongor.Small secreted membrane vesicles called exosomes have recently attracted a great interest after the discovery that they transfer mRNA that can be translated into protein in recipient cells. Surprisingly, we found that for the majority of exosomal mRNAs only a fraction of their corresponding probes is detectable on the expression microarrays. Exosomal mRNA fragmentation is characterized with a specific structural pattern. The closer to the 3′-end of the transcript the fragments are localized, the larger fraction among the secreted RNAs they constitute. Since the 3′-ends of transcripts contain elements conferring subcellular localization of mRNA and are rich in miRNA-binding sites, exosomal RNA may act as competing RNA to regulate stability, localization and translation activity of mRNAs in recipient cells. © 2013 Batagov and Kurochkin; licensee BioMed Central Ltd.