Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

This article describes the software suite GROMACS (Groningen MAchine for Chemical Simulation) that was developed at the University of Groningen, The Netherlands, in the early 1990s. The software, written in ANSI C, originates from a parallel hardware project, and is well suited for parallelization on processor clusters. By careful optimization of neighbor searching and of inner loop performance, GROMACS is a very fast program for molecular dynamics simulation. It does not have a force field of its own, but is compatible with GROMOS, OPLS, AMBER, and ENCAD force fields. In addition, it can handle polarizable shell models and flexible constraints. The program is versatile, as force routines can be added by the user, tabulated functions can be specified, and analyses can be easily customized. Nonequilibrium dynamics and free energy determinations are incorporated. Interfaces with popular quantum-chemical packages (MOPAC, GAMES-UK, GAUSSIAN) are provided to perform mixed MM/QM simulations. The package includes about 100 utility and analysis programs. GROMACS is in the public domain and distributed (with source code and documentation) under the GNU General Public License. It is maintained by a group of developers from the Universities of Groningen, Uppsala, and Stockholm, and the Max Planck Institute for Polymer Research in Mainz. Its Web site is http://www.gromacs.org.

/pdf/gromacs-fast-flexible-and-free-15aazpel5u.pdf

GROMACS: Fast, flexible, and free

Functional analysis of large gene lists, derived in most cases from emerging high-throughput genomic, proteomic and bioinformatics scanning approaches, is still a challenging and daunting task. The gene-annotation enrichment analysis is a promising high-throughput strategy that increases the likelihood for investigators to identify biological processes most pertinent to their study. Approximately 68 bioinformatics enrichment tools that are currently available in the community are collected in this survey. Tools are uniquely categorized into three major classes, according to their underlying enrichment algorithms. The comprehensive collections, unique tool classifications and associated questions/issues will provide a more comprehensive and up-to-date view regarding the advantages, pitfalls and recent trends in a simpler tool-class level rather than by a tool-by-tool approach. Thus, the survey will help tool designers/developers and experienced end users understand the underlying algorithms and pertinent details of particular tool categories/tools, enabling them to make the best choices for their particular research interests.

/pdf/bioinformatics-enrichment-tools-paths-toward-the-2yo1xl3r2j.pdf

Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists

The cBioPortal for Cancer Genomics (http://cbioportal.org) provides a Web resource for exploring, visualizing, and analyzing multidimensional cancer genomics data. The portal reduces molecular profiling data from cancer tissues and cell lines into readily understandable genetic, epigenetic, gene expression, and proteomic events. The query interface combined with customized data storage enables researchers to interactively explore genetic alterations across samples, genes, and pathways and, when available in the underlying data, to link these to clinical outcomes. The portal provides graphical summaries of gene-level data from multiple platforms, network visualization and analysis, survival analysis, patient-centric queries, and software programmatic access. The intuitive Web interface of the portal makes complex cancer genomics profiles accessible to researchers and clinicians without requiring bioinformatics expertise, thus facilitating biological discoveries. Here, we provide a practical guide to the analysis and visualization features of the cBioPortal for Cancer Genomics.

Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

The protein property prediction and testing database (PPT-DB) is a database housing nearly 30 carefully curated databases, each of which contains commonly predicted protein property information. These properties include both structural (i.e. secondary structure, contact order, disulfide pairing) and dynamic (i.e. order parameters, B-factors, folding rates) features that have been measured, derived or tabulated from a variety of sources. PPT-DB is designed to serve two purposes. First it is intended to serve as a centralized, up-to-date, freely downloadable and easily queried repository of predictable or ‘derived’ protein property data. In this role, PPT-DB can serve as a one-stop, fully standardized repository for developers to obtain the required training, testing and validation data needed for almost any kind of protein property prediction program they may wish to create. The second role that PPT-DB can play is as a tool for homology-based protein property prediction. Users may query PPT-DB with a sequence of interest and have a specific property predicted using a sequence similarity search against PPT-DB's extensive collection of proteins with known properties. PPT-DB exploits the well-known fact that protein structure and dynamic properties are highly conserved between homologous proteins. Predictions derived from PPT-DB's similarity searches are typically 85–95% correct (for categorical predictions, such as secondary structure) or exhibit correlations of >0.80 (for numeric predictions, such as accessible surface area). This performance is 10–20% better than what is typically obtained from standard ‘ab initio’ predictions. PPT-DB, its prediction utilities and all of its contents are available at http://www.pptdb.ca

PPT-DB: the protein property prediction and testing database

A relationship between ulcerative colitis (UC) and diet has been shown in epidemiological and experimental studies. In a 6-month, open-label, randomized, placebo-controlled trial, adult UC patients in clinical remission were randomized to either an “Anti-inflammatory Diet (AID)” or “Canada’s Food Guide (CFG)”. Menu plans in the AID were designed to increase the dietary intake of dietary fiber, probiotics, antioxidants, and omega-3 fatty acids and to decrease the intake of red meat, processed meat, and added sugar. Stool was collected for fecal calprotectin (FCP) and microbial analysis. Metabolomic analysis was performed on urine, serum, and stool samples at the baseline and study endpoint. In this study, 53 patients were randomized. Five (19.2%) patients in the AID and 8 (29.6%) patients in the CFG experienced a clinical relapse. The subclinical response to the intervention (defined as FCP < 150 µg/g at the endpoint) was significantly higher in the AID group (69.2 vs. 37.0%, p = 0.02). The patients in the AID group had an increased intake of zinc, phosphorus, selenium, yogurt, and seafood versus the control group. Adherence to the AID was associated with significant changes in the metabolome, with decreased fecal acetone and xanthine levels along with increased fecal taurine and urinary carnosine and p-hydroxybenzoic acid levels. The AID subjects also had increases in fecal Bifidobacteriaceae, Lachnospiraceae, and Ruminococcaceae. In this study, we found thatdietary modifications involving the increased intake of anti-inflammatory foods combined with a decreased intake of pro-inflammatory foods were associated with metabolic and microbial changes in UC patients in clinical remission and were effective in preventing subclinical inflammation.

/pdf/anti-inflammatory-diet-prevents-subclinical-colonic-2g76ylqx.pdf

Anti-Inflammatory Diet Prevents Subclinical Colonic Inflammation and Alters Metabolomic Profile of Ulcerative Colitis Patients in Clinical Remission

Metabolomics in Humans and Other Mammals

We fused various polypeptide extensions to the C-termini of single chain Fv (scFv) and disulfide-stabilized Fv (dsFv) fragments to facilitate detection of bi-functional proteins or to add biological effector domains, which included the human metallothionein (HMT) motif and biotin mimetic sequence. These bi-functional proteins were expressed and secreted in a recombinant Pichia pastoris system and showed specific anti-idiotype binding activity, as determined by competitive radioimmunoassaying. However, the fusion protein constructed with dsFv- HMT, but not scFv-HMT, had lost this binding activity. The interruption of the structural conformation as a result in dsFv-HMT may be explained by the interactions between the cysteines engineered in dsFv domains and the cysteines in the HMT region.

Construction and expression of bi-functional proteins of single-chain Fv with effector domains.

As an alternative to the conventional solvation models, we propose to use the molecular theory of solvation also known as the three-dimensional reference interaction site model with the Kovalenko–Hirata closure (3D-RISM-KH) as an essential part of a multiscale approach for biomolecular modelling at nanoscale. Based on the rigorous statistical–mechanical foundation, this approach provides a natural link between different levels of coarse-graining details in the multiscale description of the solvation structure and thermodynamics, from highly localised structural solvent and bound ligand molecules to effective desolvation potentials and self-assembling nanoarchitectures. Using the 3D-RISM-KH approach, we study all stages of formation of fibrillar aggregates and amyloid fibrils. We also show that the approach is capable of predicting binding sites for the inhibitors of the pathological conversion and aggregation of prion proteins in agreement with experimental data.

David S. Wishart

Papers

PPT-DB: the protein property prediction and testing database

Anti-Inflammatory Diet Prevents Subclinical Colonic Inflammation and Alters Metabolomic Profile of Ulcerative Colitis Patients in Clinical Remission

Metabolomics in Humans and Other Mammals

Construction and expression of bi-functional proteins of single-chain Fv with effector domains.

3D-RISM-KH approach for biomolecular modelling at nanoscale: thermodynamics of fibril formation and beyond