A5005013902- Bioinformatics and Genomic Networks
- Machine Learning in Bioinformatics
- RNA Research and Splicing
- RNA modifications and cancer
- RNA and protein synthesis mechanisms
- Protein Structure and Dynamics
- Gene expression and cancer classification
- Computational Drug Discovery Methods
- Genomics and Phylogenetic Studies
- Genomics and Chromatin Dynamics
- Biomedical Text Mining and Ontologies
- Advanced Clustering Algorithms Research
- Face and Expression Recognition
- Integrated Circuits and Semiconductor Failure Analysis
- Advancements in Photolithography Techniques
- Particle accelerators and beam dynamics
- Computability, Logic, AI Algorithms
- Cellular Automata and Applications
- Prion Diseases and Protein Misfolding
- Gene Regulatory Network Analysis
- Evolutionary Algorithms and Applications
- Machine Learning in Materials Science
- Theoretical and Computational Physics
- Machine Learning and Algorithms
- Neurological diseases and metabolism
Colorado State University
2016-2025
Sir Charles Gairdner Hospital
2023
University of Washington
2004-2008
École Nationale Supérieure des Mines de Paris
2007
Stanford University
2001-2006
Bradley University
2006
Howard Hughes Medical Institute
2005
University of Zurich
2005
Technion – Israel Institute of Technology
1999-2003
Hebrew University of Jerusalem
1996
We present a novel clustering method using the approach of support vector machines. Data points are mapped by means Gaussian kernel to high dimensional feature space, where we search for minimal enclosing sphere. This sphere, when back data can separate into several components, each cluster points. simple algorithm identifying these clusters. The width controls scale at which is probed while soft margin constant helps coping with outliers and overlapping structure dataset explored varying...
Automated annotation of protein function is challenging. As the number sequenced genomes rapidly grows, overwhelming majority products can only be annotated computationally. If computational predictions are to relied upon, it crucial that accuracy these methods high. Here we report results from first large-scale community-based critical assessment (CAFA) experiment. Fifty-four representing state art for prediction were evaluated on a target set 866 proteins 11 organisms. Two findings stand...
Motivation: Despite advances in high-throughput methods for discovering protein–protein interactions, the interaction networks of even well-studied model organisms are sketchy at best, highlighting continued need computational to help direct experimentalists search novel interactions.
Abstract Alternative splicing and alternative polyadenylation (APA) of pre-mRNAs greatly contribute to transcriptome diversity, coding capacity a genome gene regulatory mechanisms in eukaryotes. Second-generation sequencing technologies have been extensively used analyse transcriptomes. However, major limitation short-read data is that it difficult accurately predict full-length splice isoforms. Here we sequenced the sorghum using Pacific Biosciences single-molecule real-time long-read...
A major bottleneck in our understanding of the molecular underpinnings life is assignment function to proteins. While experiments provide most reliable annotation proteins, their relatively low throughput and restricted purview have led an increasing role for computational prediction. However, assessing methods protein prediction tracking progress field remain challenging.We conducted second critical assessment functional (CAFA), a timed challenge assess that automatically assign function....
Metabolomic data are frequently acquired using chromatographically coupled mass spectrometry (MS) platforms. For such datasets, the first step in analysis relies on feature detection, where a is defined by and retention time. While typically derived from single compound, spectrum of signals more more-accurate representation spectrometric signal for given metabolite. Here, we report novel grouping method that operates an unsupervised manner to group MS into spectra without relying...
Deep learning architectures have recently demonstrated their power in predicting DNA- and RNA-binding specificity. Existing methods fall into three classes: Some are based on convolutional neural networks (CNNs), others use recurrent (RNNs) rely hybrid combining CNNs RNNs. However, existing studies the relative merit of various remains unclear.In this study we present a systematic exploration deep for For purpose, deepRAM, an end-to-end tool that provides implementation wide selection...
Abstract Microbial breakdown of organic matter is one the most important processes on Earth, yet controls decomposition are poorly understood. Here we track 36 terrestrial human cadavers in three locations and show that a phylogenetically distinct, interdomain microbial network assembles during despite selection effects location, climate season. We generated metagenome-assembled genome library from cadaver-associated soils integrated it with metabolomics data to identify links between...
We carried out a large-scale screen to identify interactions between integral membrane proteins of Saccharomyces cerevisiae by using modified split-ubiquitin technique. Among 705 annotated as membrane, we identified 1,985 putative involving 536 proteins. To ascribe confidence levels the interactions, used support vector machine algorithm classify based on assay results and protein data derived from literature. Previously computationally supported were train machine, which 131 highest...
Understanding the genetic basis of HIV-1 drug resistance is essential to developing new antiretroviral drugs and optimizing use existing drugs. This understanding, however, hampered by large numbers mutation patterns associated with cross-resistance within each class. We used five statistical learning methods (decision trees, neural networks, support vector regression, least-squares least angle regression) relate protease reverse transcriptase mutations in vitro susceptibility 16 Learning...
Abstract The protein-protein interaction networks of even well-studied model organisms are sketchy at best, highlighting the continued need for computational methods to help direct experimentalists in search novel interactions. This has prompted development a number predicting interactions based on various sources data and methodologies. common method choosing negative examples training predictor is annotations cellular localization, observation that pairs proteins have different...
Clustering is one of the most commonly used tools in analysis gene expression data (1,2). The usage grouping genes based on premise that coexpression a result coregulation. It often as preliminary step extracting networks and inference function (3,4). experiments can be to discover novel phenotypic aspects cells tissues (3,5,6), including sensitivity drugs (7), also detect artifacts experimental conditions (8). its applications biology are presented greater detail Chapter 13 (see ref. 9)....
Remote homology detection is the problem of detecting in cases low sequence similarity. It a hard computational with no approach that works well all cases.We present method for remote based on presence discrete motifs. The motif content pair sequences used to define similarity as kernel Support Vector Machine (SVM) classifier. We test two tasks: prediction previously unseen SCOP family and an enzyme class given other enzymes have similar function substrates. find it performs significantly...
We propose a method for predicting splice graphs that enhances curated gene models using evidence from RNA-Seq and EST alignments. Results obtained experiments in Arabidopsis thaliana show predictions made by our SpliceGrapher are more consistent with current than TAU Cufflinks. Furthermore, analysis of plant human data indicates the machine learning approach used is useful discriminating between real spurious sites, can improve reliability detection alternative splicing. available download...
Prions are important disease agents and epigenetic regulatory elements. Prion formation involves the structural conversion of proteins from a soluble form into an insoluble amyloid form. In many cases, this is driven by glutamine/asparagine (Q/N)-rich prion-forming domain. However, our understanding sequence requirements for prion propagation Q/N-rich domains has been insufficient accurate propensity prediction or domain design. By focusing exclusively on amino acid composition, we have...
We present a novel partner-specific protein-protein interaction site prediction method called PAIRpred. Unlike most existing machine learning binding methods, PAIRpred uses information from both proteins in protein complex to predict pairs of interacting residues the two proteins. captures sequence and structure about residue through pairwise kernels that are used for training support vector classifier. As result, presents more detailed model binding, offers state art accuracy predicting...
Drought is a major limiting factor of crop yields. In response to drought, plants reprogram their gene expression, which ultimately regulates multitude biochemical and physiological processes. The timing this reprogramming the nature drought-regulated genes in different genotypes are thought confer differential tolerance drought stress. Sorghum highly drought-tolerant has been increasingly used as model cereal identify that tolerance. Also, there considerable natural variation resistance...
Abiotic stresses affect plant physiology, development, growth, and alter pre-mRNA splicing. Western poplar is a model woody tree potential bioenergy feedstock. To investigate the extent of stress-regulated alternative splicing, we conducted an in-depth survey leaf, root, stem xylem transcriptomes under drought, salt, or temperature stress. Analysis approximately one billion genome-aligned RNA-Seq reads from tissue- stress-specific libraries revealed over fifteen millions novel splice...
Plant SR45 and its metazoan ortholog RNPS1 are serine/arginine-rich (SR)-like RNA binding proteins that function in splicing/postsplicing events regulate diverse processes eukaryotes. Interactions of with both RNAs crucial for regulating processing. However, vivo targets currently unclear. Using immunoprecipitation followed by high-throughput sequencing, we identified over 4000 Arabidopsis thaliana directly or indirectly associate SR45, designated as SR45-associated (SARs). Comprehensive...
Summary Comprehensive, automated software testing requires an oracle to check whether the output produced by a test case matches expected behaviour of programme. But challenges in creating suitable oracles limit ability perform some programmes, and especially scientific software. Metamorphic is method for automating process programmes without oracles. This technique operates checking programme behaves according properties called metamorphic relations . A relation describes change when input...