Aim A Neural network (NN) to predict serologic specificity of HLA alleles was first introduced in 2003 by Maiers et al. The procedure involved several manual steps and prediction errors were estimated from an arbitrar...
Aim A Neural network (NN) to predict serologic specificity of HLA alleles was first introduced in 2003 by Maiers et al. The procedure involved several manual steps and prediction errors were estimated from an arbitrarily selected single partition of available labeled data. However, these might not reflect the true predictive capability of the model. Our aim was to provide an automated framework for the existing NN, estimate the true prediction accuracy and capability using a cross-validation model and compare with other predictive machine learning techniques. Methods An NN framework was created and validated in the R software suite to predict serologic specificities for HLA-A, B, C, DQB1, DRB1 and DPB1. We established a method to tune NN parameters using a grid technique with leave one out cross validation and compared the prediction accuracy of the NN with a k-Nearest Neighbors (kNN) technique. Results A concordance between the manual and automated NN of upwards of 96% was obtained for all loci, measured on an independent unlabelled test dataset. With parameter tuning techniques for the NN we have estimated the average prediction accuracy across all loci to be 86% .We also observed that kNN provide d better prediction accuracy of 90% (average). These numbers were slightly lower than the previous NN validation based on a single partition of the data, but still provide acceptable performance for both the NN and kNN. Conclusions Human involvement in NN model selection is both extremely time consuming and prone to errors. An automated framework minimizes the risks of over-training by monitoring variations in errors. Cross-validation techniques are known to provide better generalization error estimates. In this context, our reported error estimates maybe closer to the actual error rates one can achieve with similar models. We also demonstrate kNN outperforms the existing NN. This result should support efforts in finding relationships between alleles and associated serol
Background: CoreGenes3.5 is a webserver that determines sets of core genes from viral and small bacterial genomes as an automated batch process. Previous versions of CoreGenes have been used to classify bacteriophage ...
详细信息
In this paper, we present an approach for extending the existing concept of nanopublications-tiny entities of scientific results in RDF representation-to broaden their application range. The proposed extension uses En...
详细信息
RECOMB 2013 was successfully held in Tsinghua University, Beijing, China on April 7-10, 2013, hosted by the bioinformatics Division and Center for Synthetic and Systems biology, Tsinghua National Laboratory for Inform...
RECOMB 2013 was successfully held in Tsinghua University, Beijing, China on April 7-10, 2013, hosted by the bioinformatics Division and Center for Synthetic and Systems biology, Tsinghua National Laboratory for Information Science and Technology （TNLIST）. A total of about 500 professionals from both academia and industry from 29 countries and regions attended the conference and its RECOMB-Seq satellite workshop after the main conference.
作者：
Ahmed, ZeeshanZeeshan, SamanHuber, ClaudiaHensel, MichaelSchomburg, DietmarMünch, RichardEylert, EvaEisenreich, WolfgangDandekar, ThomasDepartment of Bioinformatics
Biocenter University of Würzburg Am Hubland Department of Neurobiology and Genetics Biocenter University of Wuerzburg Am Hubland Institute of Molecular and Translational Therapeutic Strategies Hannover Medical School Lehrstuhl für Biochemie Center of Isotopologue Profiling Technische Universität München Division of Microbiology University of Osnabrück Department of Bioinformatics and Biochemistry Technical University Braunschweig Institute for Microbiology Biozentrum Technical University Braunschweig Germany and Computational biology and structures program European Molecular Biology Laboratory Germany Department of Bioinformatics Biocenter University of Würzburg Am Hubland Department of Neurobiology and Genetics Biocenter University of Wuerzburg Am Hubland Institute of Molecular and Translational Therapeutic Strategies Hannover Medical School Lehrstuhl für Biochemie Center of Isotopologue Profiling Technische Universität München Division of Microbiology University of Osnabrück Department of Bioinformatics and Biochemistry Technical University Braunschweig Institute for Microbiology Biozentrum Technical University Braunschweig Germany and Computational biology and structures program European Molecular Biology Laboratory 97074 Wuerzburg 97074 Wuerzburg Carl-Neuberg-Str. 1 Lichtenbergstraße 4 Barbarastraße 11 Gebäude 36 49076 Osnabrück Langer Kamp 19B 2. Obergeschoss Spielmannstraße 7 Meyerhofstr. 1 97074 Wuerzburg 97074 Wuerzburg Carl-Neuberg-Str. 1 Lichtenbergstraße 4 Barbarastraße 11 Gebäude 36 49076 Osnabrück Langer Kamp 19B 2. Obergeschoss Spielmannstraße 7 Braunschweig 38106 Hanover Germany Department of Bioinformatics
Biocenter University of Würzburg Am Hubland Department of Neurobiology and Genetics Biocenter University of Wuerzburg Am Hubland Institute of Molecular and Translational Therapeutic Strategies Hannover Medical School Lehrstuhl für Biochemie Center of Isotopologue Profiling Technische Univer
UNLABELLED: The composition of stable-isotope labelled isotopologues/isotopomers in metabolic products can be measured by mass spectrometry and supports the analysis of pathways and fluxes. As a prerequisite, the orig...
Background: Understanding crosstalk and feedback among oncogenic pathways is a critical challenge to overcoming resistance to targeted anticancer therapy. The topology of biological networks has increasingly been used...
详细信息
ISBN:
(纸本)9781622769711
Background: Understanding crosstalk and feedback among oncogenic pathways is a critical challenge to overcoming resistance to targeted anticancer therapy. The topology of biological networks has increasingly been used to complement studies based on individual genes or gene collections. We apply network analysis to understand interactions among critical biological signaling pathways in breast cancer. Results: An overlapping clustering of the networks showed highly distinct patterns in patients with high IGF versus low IGF axis. We demonstrate through cluster comparison metrics and permutation studies that these differences are unlikely to occur by chance. IRS-1, an IGF adaptor protein, is shown to have a highly central place in the IGF high as compared to the IGF low networks. We further demonstrate that network connections reveal information about interactions of genes in TGF-beta, MAPK, and other pathways known to interact with IGF. Conclusions: Network analyses can provide novel insights and hypotheses about signaling pathways involved in feedback and crosstalk in breast cancer.
Asymmetric patchy particle models have recently been shown to describe the crystallization of small globular proteins with near-quantitative accuracy. Here, we investigate how asymmetry in patch geometry and bond ener...
详细信息
Asymmetric patchy particle models have recently been shown to describe the crystallization of small globular proteins with near-quantitative accuracy. Here, we investigate how asymmetry in patch geometry and bond energy generally impacts the phase diagram and nucleation dynamics of this family of soft matter models. We find the role of the geometry asymmetry to be weak, but the energy asymmetry to markedly interfere with the crystallization thermodynamics and kinetics. These results provide a rationale for the success and occasional failure of the proposal of George and Wilson for protein crystallization conditions as well as physical guidance for developing more effective protein crystallization strategies.
Understanding the geometric, topologic, and mechanical properties of cells and their interactions is critical for studying tissue pattern formation and organ development. computational model and tools for simulating c...
详细信息
ISBN:
(纸本)9781457702150
Understanding the geometric, topologic, and mechanical properties of cells and their interactions is critical for studying tissue pattern formation and organ development. computational model and tools for simulating cell pattern formation have broad implications in studying embryogenesis, blood-vessel development, tissue regeneration, and tumor growth. Although a number of cell modeling methods exist, they do not simultaneously account for detailed cellular shapes as well as dynamic changes in cell geometry and topology. Here we describe a dynamic finite element cell model (dFEMC) for studying populations of cells and tissue development. By incorporating details of cell shape, cell growth and shrinkage, cell birth and death, cell division and fusion, our method can model realistically a variety problems of cell pattern formation. We give two examples of applying our method to the study of cell fusion and cell apoptosis. The dFEMC model developed here provides a general computational framework for studying dynamics pattern formation of tissue.
One known challenge in analyzing gene expression data is to combine analysis outcomes obtained disparately by applying multiple, independent meta-analysis methods. Here we present an integrative computational system t...
详细信息
One known challenge in analyzing gene expression data is to combine analysis outcomes obtained disparately by applying multiple, independent meta-analysis methods. Here we present an integrative computational system that narrows down biological hypotheses by integrating gene expression patterns, transcription factor (TF) binding site analysis outcomes, and Gene Ontology (GO) enrichment analysis outcomes. This system identifies regulated genes from microarray experiments through statistical processes, categorizes similarly behaving groups of genes and then carries out binding site analysis and gene function enrichment analysis based on some significant clusters. The output is an ordered set of "putative" pair-wise relationships between TFs and their potential target genes. The relationships are ranked based on their closeness to the experimental context. We demonstrate the effectiveness of our framework using two independent microarray data sets.
We combined single-molecule force spectroscopy with nuclear magnetic resonance measurements and molecular mechanics simulations to examine overstretching transitions in single-stranded nucleic acids. In single-strande...
详细信息
We combined single-molecule force spectroscopy with nuclear magnetic resonance measurements and molecular mechanics simulations to examine overstretching transitions in single-stranded nucleic acids. In single-stranded DNA and single-stranded RNA there is a low-force transition that involves unwinding of the helical structure, along with base unstacking. We determined that the high-force transition that occurs in polydeoxyadenylic acid single-stranded DNA is caused by the cooperative forced flipping of the dihedral angle formed between four atoms, O5’-C5’-C4’-C3’ (γ torsion), in the nucleic acid backbone within the canonical B-type helix. The γ torsion also flips under force in A-type helices, where the helix is shorter and wider as compared to the B-type helix, but this transition is less cooperative than in the B type and does not generate a high-force plateau in the force spectrums of A-type helices. We find that a similar high-force transition can be induced in polyadenylic acid single-stranded RNA by urea, presumably due to disrupting the intramolecular hydrogen bonding in the backbone. We hypothesize that a pronounced high-force transition observed for B-type helices of double stranded DNA also involves a cooperative flip of the γ torsion. These observations suggest new fundamental relationships between the canonical structures of single-and double-stranded DNA and the mechanism of their molecular elasticity.
暂无评论