Thorsteinn Rögnvaldsson

JETNET 3.0—A versatile artificial neural network package

An F77 package for feed-forward artificial neural network data processing, JETNET 3.0, is presented. It represents a substantial extension and generalization of an earlier release, JETNET 2.0. The package, which consists of a set of subroutines, is focused on multilayer perceptron architectures. As compared to earlier versions it contains a variety of minimization options, measures for monitoring the learning process, limited precision emulation, etc. Also, the reader is provided with a set of guidelines for when to use the different options.

Why neural networks should not be used for HIV-1 protease cleavage site prediction

UNLABELLED Several papers have been published where nonlinear machine learning algorithms, e.g. artificial neural networks, support vector machines and decision trees, have been used to model the specificity of the HIV-1 protease and extract specificity rules. We show that the dataset used in these studies is linearly separable and that it is a misuse of nonlinear classifiers to apply them to this problem. The best solution on this dataset is achieved using a linear classifier like the simple perceptron or the linear support vector machine, and it is straightforward to extract rules from these linear models. We identify key residues in peptides that are efficiently cleaved by the HIV-1 protease and list the most prominent rules, relating them to experimental results for the HIV-1 protease. MOTIVATION Understanding HIV-1 protease specificity is important when designing HIV inhibitors and several different machine learning algorithms have been applied to the problem. However, little progress has been made in understanding the specificity because nonlinear and overly complex models have been used. RESULTS We show that the problem is much easier than what has previously been reported and that linear classifiers like the simple perceptron or linear support vector machines are at least as good predictors as nonlinear algorithms. We also show how sets of specificity rules can be generated from the resulting linear classifiers. AVAILABILITY The datasets used are available at http://www.hh.se/staff/bioinf/

Modular, scriptable and automated analysis tools for high-throughput peptide mass fingerprinting

UNLABELLED A set of new algorithms and software tools for automatic protein identification using peptide mass fingerprinting is presented. The software is automatic, fast and modular to suit different laboratory needs, and it can be operated either via a Java user interface or called from within scripts. The software modules do peak extraction, peak filtering and protein database matching, and communicate via XML. Individual modules can therefore easily be replaced with other software if desired, and all intermediate results are available to the user. The algorithms are designed to operate without human intervention and contain several novel approaches. The performance and capabilities of the software is illustrated on spectra from different mass spectrometer manufacturers, and the factors influencing successful identification are discussed and quantified. MOTIVATION Protein identification with mass spectrometric methods is a key step in modern proteomics studies. Some tools are available today for doing different steps in the analysis. Only a few commercial systems integrate all the steps in the analysis, often for only one vendors hardware, and the details of these systems are not public. RESULTS A complete system for doing protein identification with peptide mass fingerprints is presented, including everything from peak picking to matching the database protein. The details of the different algorithms are disclosed so that academic researchers can have full control of their tools. AVAILABILITY The described software tools are available from the Halmstad University website www.hh.se/staff/bioinf/ SUPPLEMENTARY INFORMATION Details of the algorithms are described in supporting information available from the Halmstad University website www.hh.se/staff/bioinf/

Using neural networks to identify jets

Abstract A neural network method for identifying the ancestor of a hadron jet is presented. The idea is to find an efficient mapping between certain observed hadronic kinematical variables and the quark-gluon identity. This is done with a neuronic expansion in terms of a network of sigmoidal functions using a gradient descent procedure, where the errors are back-propagated through the network. With this method we are able to separate gluon from quark jets originating from Monte Carlo generated e + e − events with ≈85% approach. The result is independent of the MC model used. This approach for isolating the gluon jet is then used to study the so-called string effect. In addition, heavy quarks (b and c) in e + e − reactions can be identified on the 50% level by just observing the hadrons. In particular we are able to separate b-quarks with an efficiency and purity, which is comparable with what is expected from vertex detectors. We also speculate on how the neutral network method can be used to disentangle different hadronization schemes by compressing the dimensionality of the state space of hadrons.

On Langevin updating in multilayer perceptrons

The Langevin updating rule, in which noise is added to the weights during learning, is presented and shown to improve learning on problems with initially ill-conditioned Hessians. This is particularly important for multilayer perceptrons with many hidden layers, that often have ill-conditioned Hessians. In addition, Manhattan updating is shown to have a similar effect.

Comprehensive Bioinformatic Analysis of the Specificity of Human Immunodeficiency Virus Type 1 Protease

ABSTRACT Rapidly developing viral resistance to licensed human immunodeficiency virus type 1 (HIV-1) protease inhibitors is an increasing problem in the treatment of HIV-infected individuals and AIDS patients. A rational design of more effective protease inhibitors and discovery of potential biological substrates for the HIV-1 protease require accurate models for protease cleavage specificity. In this study, several popular bioinformatic machine learning methods, including support vector machines and artificial neural networks, were used to analyze the specificity of the HIV-1 protease. A new, extensive data set (746 peptides that have been experimentally tested for cleavage by the HIV-1 protease) was compiled, and the data were used to construct different classifiers that predicted whether the protease would cleave a given peptide substrate or not. The best predictor was a nonlinear predictor using two physicochemical parameters (hydrophobicity, or alternatively polarity, and size) for the amino acids, indicating that these properties are the key features recognized by the HIV-1 protease. The present in silico study provides new and important insights into the workings of the HIV-1 protease at the molecular level, supporting the recent hypothesis that the protease primarily recognizes a conformation rather than a specific amino acid sequence. Furthermore, we demonstrate that the presence of 1 to 2 lysine residues near the cleavage site of octameric peptide substrates seems to prevent cleavage efficiently, suggesting that this positively charged amino acid plays an important role in hindering the activity of the HIV-1 protease.

Robust AFR Estimation Using the Ion Current and Neural Networks

A robust air/fuel ratio “soft sensor” is presented based on non-linear signal processing of the ion current signal using neural networks. Care is taken to make the system insensitive to amplitude variations, due to e.g. fuel additives, by suitable preprocessing of the signal.

Spark Advance Control Using the Ion Current and Neural Soft Sensors

Two spark advance control systems are outlined; both based on feedback from nonlinear neural network soft sensors and ion current detection. One uses an estimate on the location of the pressure peak and the other uses an estimate of the location of the center of combustion. Both quantities are estimated from the ion current signal using neural networks. The estimates are correct within roughly two crank angle degrees when evaluated on a cycle to cycle basis, and roughly within one crank angle degree when the quantities are averaged over consecutive cycles.The pressure peak detection based control system is demonstrated on a SAAB 9000 car, equipped with a 2.3 liter low-pressure turbo charged engine, during normal highway driving.

Recognition of similarities (ROS): A methodological approach to analysing and characterising patterns of daily occupations

Abstract It has been proposed that it should be possible to identify patterns of daily occupations that promote health or cause illness. This study aimed to develop and to evaluate.a process for analysing and characterising subjectively perceived patterns of daily occupations, by describing patterns as consisting of main, hidden, and unexpected occupations. Yesterday diaries describing one day of 100 working married mothers were collected through interviews. The diaries were transformed into time‐and‐occupation graphs. An analysis based on visual interpretation of the patterns was performed. The graphs were grouped into the categories low, medium, or high complexity. In order to identify similarities the graphs were then compared both pair‐wise and group‐wise. Finally, the complexity and similarities perspectives were integrated, identifying the most typical patterns of daily occupations representing low, medium, and high complexity. Visual differences in complexity were evident. In order to validate the Recognition of Similarities (ROS) process developed, a measure expressing the probability of change was computed. This probability was found to differ statistically significantly between the three groups, supporting the validity of the ROS process.

How to find simple and accurate rules for viral protease cleavage specificities

BackgroundProteases of human pathogens are becoming increasingly important drug targets, hence it is necessary to understand their substrate specificity and to interpret this knowledge in practically useful ways. New methods are being developed that produce large amounts of cleavage information for individual proteases and some have been applied to extract cleavage rules from data. However, the hitherto proposed methods for extracting rules have been neither easy to understand nor very accurate. To be practically useful, cleavage rules should be accurate, compact, and expressed in an easily understandable way.ResultsA new method is presented for producing cleavage rules for viral proteases with seemingly complex cleavage profiles. The method is based on orthogonal search-based rule extraction (OSRE) combined with spectral clustering. It is demonstrated on substrate data sets for human immunodeficiency virus type 1 (HIV-1) protease and hepatitis C (HCV) NS3/4A protease, showing excellent prediction performance for both HIV-1 cleavage and HCV NS3/4A cleavage, agreeing with observed HCV genotype differences. New cleavage rules (consensus sequences) are suggested for HIV-1 and HCV NS3/4A cleavages. The practical usability of the method is also demonstrated by using it to predict the location of an internal cleavage site in the HCV NS3 protease and to correct the location of a previously reported internal cleavage site in the HCV NS3 protease. The method is fast to converge and yields accurate rules, on par with previous results for HIV-1 protease and better than previous state-of-the-art for HCV NS3/4A protease. Moreover, the rules are fewer and simpler than previously obtained with rule extraction methods.ConclusionA rule extraction methodology by searching for multivariate low-order predicates yields results that significantly outperform existing rule bases on out-of-sample data, but are more transparent to expert users. The approach yields rules that are easy to use and useful for interpreting experimental data.