Artificial neural networks: fundamentals, computing, design, and application

Artificial neural networks (ANNs) are relatively new computational tools that have found extensive utilization in solving many complex real-world problems. The attractiveness of ANNs comes from their remarkable information processing characteristics pertinent mainly to nonlinearity, high parallelism, fault and noise tolerance, and learning and generalization capabilities. This paper aims to familiarize the reader with ANN-based computing (neurocomputing) and to serve as a useful companion practical guide and toolkit for the ANNs modeler along the course of ANN project development. The history of the evolution of neurocomputing and its relation to the field of neurobiology is briefly discussed. ANNs are compared to both expert systems and statistical regression and their advantages and limitations are outlined. A bird's eye review of the various types of ANNs and the related learning rules is presented, with special emphasis on backpropagation (BP) ANNs theory and design. A generalized methodology for developing successful ANNs projects from conceptualization, to design, to implementation, is described. The most common problems that BPANNs developers face during training are summarized in conjunction with possible causes and remedies. Finally, as a practical application, BPANNs were used to model the microbial growth curves of S. flexneri. The developed model was reasonably accurate in simulating both training and test time-dependent growth curves as affected by temperature and pH.     

Biocatalysts: measurement, modelling and design of heterogeneity

Multiple phenomena are involved in conversions by immobilized biocatalysts. A paradox is identified between analytical desires on one hand and analytical boundary conditions on the other: while the study of interdependent phenomena would call for their simultaneous analysis in an integrated context, the available experimental options may impose a series of separate and dedicated analyses. From this analysis, bottlenecks in particle performance may be identified, if possible supported by a mechanistic model and performance criteria. Subsequently, a strategy for further biocatalyst development may be chosen. Finally, possibilities for future improvement of biocatalysts are discussed for various fields of research. Some examples of recent developments in enzyme and matrix characteristics, reactor operation, and micro-technology are discussed.     

Yabi: An online research environment for grid,high performance and cloud computing

Background  

There is a significant demand for creating pipelines or workflows in the life science discipline that chain a number of discrete compute and data intensive analysis tasks into sophisticated analysis procedures. This need has led to the development of general as well as domain-specific workflow environments that are either complex desktop applications or Internet-based applications. Complexities can arise when configuring these applications in heterogeneous compute and storage environments if the execution and data access models are not designed appropriately. These complexities manifest themselves through limited access to available HPC resources, significant overhead required to configure tools and inability for users to simply manage files across heterogenous HPC storage infrastructure.     

A simplified method for power-law modelling of metabolic pathways from time-course data and steady-state flux profiles

Background  

In order to improve understanding of metabolic systems there have been attempts to construct S-system models from time courses. Conventionally, non-linear curve-fitting algorithms have been used for modelling, because of the non-linear properties of parameter estimation from time series. However, the huge iterative calculations required have hindered the development of large-scale metabolic pathway models. To solve this problem we propose a novel method involving power-law modelling of metabolic pathways from the Jacobian of the targeted system and the steady-state flux profiles by linearization of S-systems.     

A logic-reasoning based system to harness bioprocess experimental data and knowledge for design

Bioprocess design requires substantial resources during the experimental investigation of the options for each bioprocess step. This is both time-consuming and expensive. The amount of data available has increased exponentially since the expansion of new biological drug development. Data are heterogeneous, sometimes inconsistent and incomplete, making them difficult to be systematically utilised for analysis for any new bioprocess design. In this paper, we report a novel computational method that harnesses the bioprocess experimental data to assist design decision making, and perhaps identify further needed experiments. First, we develop a new data representation structure to capture the experimental data systematically. Then the ontology for modelling the relationship of data properties is created. A computational system has been developed to search relevant data, or to predict required process conditions, or to suggest a new set of experiments for process development. A prototype for harnessing centrifugation experimental data has been built, and is then used to illustrate the method and demonstrate the type of results that can be obtained. Evaluations show that such a system has significant potential to mine the relevant experimental data to assist new drug bioprocess development, which should reduce process development time and cost.     

Minimum Information about a Cardiac Electrophysiology Experiment (MICEE): standardised reporting for model reproducibility, interoperability, and data sharing

Cardiac experimental electrophysiology is in need of a well-defined Minimum Information Standard for recording, annotating, and reporting experimental data. As a step towards establishing this, we present a draft standard, called Minimum Information about a Cardiac Electrophysiology Experiment (MICEE). The ultimate goal is to develop a useful tool for cardiac electrophysiologists which facilitates and improves dissemination of the minimum information necessary for reproduction of cardiac electrophysiology research, allowing for easier comparison and utilisation of findings by others. It is hoped that this will enhance the integration of individual results into experimental, computational, and conceptual models. In its present form, this draft is intended for assessment and development by the research community. We invite the reader to join this effort, and, if deemed productive, implement the Minimum Information about a Cardiac Electrophysiology Experiment standard in their own work.     

A simple method for computing sprint acceleration kinetics from running velocity data: Replication study with improved design

Measuring the ground reaction forces (GRF) underlying sprint acceleration is important to understanding the performance of such a common task. Until recently direct measurements of GRF during sprinting were limited to a few steps per trial, but a simple method (SM) was developed to estimate GRF across an entire acceleration. The SM utilizes displacement- or velocity-time data and basic computations applied to the runner’s center of mass and was validated against compiled force plate (FP) measurements; however, this validation used multiple-trials to generate a single acceleration profile, and consequently fatigue and error may have introduced noise into the analyses. In this study, we replicated the original validation by comparing the main sprint kinetics and force-velocity-power variables (e.g. GRF and its horizontal and vertical components, mechanical power output, ratio of horizontal component to resultant GRF) between synchronized FP data from a single sprinting acceleration and SM data derived from running velocity measured with a 100 Hz laser. These analyses were made possible thanks to a newly developed 50-m FP system providing seamless GRF data during a single sprint acceleration. Sixteen trained male sprinters performed two all-out 60-m sprints. We observed good agreement between the two methods for kinetic variables (e.g. grand average bias of 4.71%, range 0.696 ± 0.540–8.26 ± 5.51%), and high inter-trial reliability (grand average standard error of measurement of 2.50% for FP and 2.36% for the SM). This replication study clearly shows that when implemented correctly, this method accurately estimates sprint acceleration kinetics.