Kert Viele

Statistics and bioinformatics in nutritional sciences: analysis of complex data in the era of systems biology.

Over the past 2 decades, there have been revolutionary developments in life science technologies characterized by high throughput, high efficiency, and rapid computation. Nutritionists now have the advanced methodologies for the analysis of DNA, RNA, protein, low-molecular-weight metabolites, as well as access to bioinformatics databases. Statistics, which can be defined as the process of making scientific inferences from data that contain variability, has historically played an integral role in advancing nutritional sciences. Currently, in the era of systems biology, statistics has become an increasingly important tool to quantitatively analyze information about biological macromolecules. This article describes general terms used in statistical analysis of large, complex experimental data. These terms include experimental design, power analysis, sample size calculation, and experimental errors (Type I and II errors) for nutritional studies at population, tissue, cellular, and molecular levels. In addition, we highlighted various sources of experimental variations in studies involving microarray gene expression, real-time polymerase chain reaction, proteomics, and other bioinformatics technologies. Moreover, we provided guidelines for nutritionists and other biomedical scientists to plan and conduct studies and to analyze the complex data. Appropriate statistical analyses are expected to make an important contribution to solving major nutrition-associated problems in humans and animals (including obesity, diabetes, cardiovascular disease, cancer, ageing, and intrauterine growth retardation).

Modeling with Mixtures of Linear Regressions

Consider data (x1,y1),...,(xn,yn), where each xi may be vector valued, and the distribution of yi given xi is a mixture of linear regressions. This provides a generalization of mixture models which do not include covariates in the mixture formulation. This mixture of linear regressions formulation has appeared in the computer science literature under the name “Hierarchical Mixtures of Experts” model.This model has been considered from both frequentist and Bayesian viewpoints. We focus on the Bayesian formulation. Previously, estimation of the mixture of linear regression model has been done through straightforward Gibbs sampling with latent variables. This paper contributes to this field in three major areas. First, we provide a theoretical underpinning to the Bayesian implementation by demonstrating consistency of the posterior distribution. This demonstration is done by extending results in Barron, Schervish and Wasserman (Annals of Statistics 27: 536–561, 1999) on bracketing entropy to the regression setting. Second, we demonstrate through examples that straightforward Gibbs sampling may fail to effectively explore the posterior distribution and provide alternative algorithms that are more accurate. Third, we demonstrate the usefulness of the mixture of linear regressions framework in Bayesian robust regression. The methods described in the paper are applied to two examples.

Influence of PCPA and MDMA (ecstasy) on physiology, development and behavior in Drosophila melanogaster

The effects of para‐chlorophenylalanine (PCPA) and 3,4 methylenedioxy‐methamphetamine (MDMA, ‘ecstasy’) were investigated in relation to development, behavior and physiology in larval Drosophila. PCPA blocks the synthesis of serotonin (5‐HT) and MDMA is known to deplete 5‐HT in mammalian neurons; thus these studies were conducted primarily to target the serotonergic system. Treatment with PCPA and MDMA delayed time to pupation and eclosion. The developmental rate was investigated with a survival analysis statistical approach that is unique for Drosophila studies. Locomotion and eating were reduced in animals exposed to MDMA or PCPA. Sensitivity to exogenously applied 5‐HT on an evoked sensory–central nervous system (CNS)–motor circuit showed that the CNS is sensitive to 5‐HT but that when depleted of 5‐HT by PCPA a decreased sensitivity occurred. A diet with MDMA produced an enhanced response to exogenous 5‐HT on the central circuit. Larvae eating MDMA from the first to third instar did not show a reduction in 5‐HT within the CNS; however, eating PCPA reduced 5‐HT as well as dopamine content as measured by high performance liquid chromatography from larval brains. As the heart serves as a good bioindex of 5‐HT exposure, it was used in larvae fed PCPA and MDMA but no significant effects occurred with exogenous 5‐HT. In summary, the action of these pharmacological compounds altered larval behaviors and development. PCPA treatment changed the sensitivity in the CNS to 5‐HT, suggesting that 5‐HT receptor regulation is modulated by neural activity of the serotonergic neurons. The actions of acute MDMA exposure suggest a 5‐HT agonist action or possible dumping of 5‐HT from neurons.

Parsimonious estimation of multiplicative interaction in analysis of variance using Kullback–Leibler Information

Abstract Many standard methods for modeling interaction in two-way ANOVA require mn interaction parameters, where m and n are the number of rows and columns in the table. By viewing the interaction parameters as a matrix and performing a singular value decomposition, one arrives at the additive main effects and multiplicative interaction (AMMI) model which is commonly used in agriculture. By using only those interaction components with the largest singular values, one can produce an estimate of interaction that requires far fewer than mn parameters while retaining most of the explanatory power of standard methods. The central inference problems of estimating the parameters and determining the number of interaction components has been difficult except in “ideal” situations (equal cell sizes, equal variance, etc.). The Bayesian methodology developed in this paper applies for unequal sample sizes and heteroscedastic data, and may be easily generalized to more complicated data structures. We illustrate the proposed methodology with two examples.

Practice patterns when treating patients with low back pain: a survey of physical therapists.

Abstract Low back pain (LBP), is a common musculoskeletal problem, affecting 75–85% of adults in their lifetime. Direct costs of LBP in the USA were estimated over 85 billion dollars in 2005 resulting in a significant economic burden for the healthcare system. LBP classification systems and outcome measures are available to guide physical therapy assessments and intervention. However, little is known about which, if any, physical therapists use in clinical practice. The purpose of this study was to identify the use of and barriers to LBP classification systems and outcome measures among physical therapists in one state. A mixed methods study using a cross-sectional cohort design with descriptive qualitative methods was performed. A survey collected both quantitative and qualitative data relevant to classification systems and outcome measures used by physical therapists working with patients with LBP. Physical therapists responded using classification systems designed to direct treatment predominantly. The McKenzie method was the most frequent approach to classify LBP. Barriers to use of classification systems and outcome measures were lack of knowledge, too limiting and time. Classification systems are being used for decision-making in physical therapy practice for patients with LBP. Lack of knowledge and training seems to be the main barrier to the use of classification systems in practice. The Oswestry Disability Index and Numerical Pain Scale were the most commonly used outcome measures. The main barrier to their use was lack of time. Continuing education and reading the literature were identified as important tools to teach evidence-based practice to physical therapists in practice.

Automated Classification of Evoked Quantal Events

We provide both theoretical and computational improvements to the analysis of synaptic transmission data. Theoretically, we demonstrate the correlation structure of observations within evoked postsynaptic potentials (EPSP) are consistent with multiple random draws from a common autoregressive moving-average (ARMA) process of order (2, 2). We use this observation and standard time series results to construct a statistical hypothesis testing procedure for determining whether a given trace is an EPSP. Computationally, we implement this method in R, a freeware statistical language, which reduces the amount of time required for the investigator to classify traces into EPSPs or non-EPSPs and eliminates investigator subjectivity from this classification. In addition, we provide a computational method for calculating common functionals of EPSPs (peak amplitude, decay rate, etc.). The methodology is freely available over the internet. The automated procedure to index the quantal characteristics greatly facilitates determining if any one or multiple parameters are changing due to experimental conditions. In our experience, the software reduces the time required to perform these analyses from hours to minutes.

Structure/Function Assessment of Synapses at Motor Nerve Terminals

The release of transmitter at neuromuscular junctions (NMJ) of the opener muscle in crayfish is quantal in nature. This NMJ offers the advantage of being able to record quantal events at specific visually identified release sites, thus allowing measurement of the physiological parameters of vesicle release and its response to be directly correlated with synaptic structure. These experiments take advantage of areas between the varicosities on the nerve terminal that we define as “stems.” Stems were chosen as the region to study because of their low synaptic output due to fewer synaptic sites. Through 3D reconstruction from hundreds of serial sections, obtained by transmission electron microscopy (TEM), at a site in which focal macropatch recordings were obtained, the number of synapses and AZs are revealed. Thus, physiological profiles with various stimulation conditions can be assessed in regards to direct synaptic structure. Here, we used the properties of the quantal shape to determine if distinct subsets of quantal signatures existed and if differences in the distributions are present depending on the frequency of stimulation. Such a quantal signature could come about by parameters of area, rise time, peak amplitude, latency, and tau decay. In this study, it is shown that even at defined sites on the stem, with few active zones, synaptic transmission is still complex and the quantal responses appear to be variable even for a given synapse over time. In this study, we could not identify a quantal signature for the conditions utilized. Synapse, 2011.

Comparative study of environmental factors influencing motor task learning and memory retention in sighted and blind crayfish

In classical conditioning, an alteration in response occurs when two stimuli are regularly paired in close succession. An area of particular research interest is classical conditioning with a chemical signal and visual and/or tactile stimuli as the unconditional stimuli, to test manipulative and motor behaviors in a learning paradigm. A classical learning task chamber was developed to examine learning trends in a sighted surface‐dwelling crayfish, Procambarus clarkii, and in a blind cave‐dwelling crayfish, Orconectes australis packardi. We examined whether learning is influenced by environmental factors and/or reliance on different primary sensory modalities. Crayfish were trained to manipulate a large, cumbersome cheliped through a small access point to obtain a food reward. In both species, acquisition of the learning task was rapid when they were in nonstressed conditions. The blind crayfish tested in low white light did not successfully complete the task, suggesting a stress response.

Right-Turn-on-Red Volume Estimation and Incremental Capacity Models for Shared Lanes at Signalized Intersections

The Highway Capacity Manual (HCM) is one of the most widely used transportation documents in the world. The signalized intersections methodology, one of the most frequently used portions of the manual, is used to estimate capacity and average control delay for individual lane groups and for intersections as a whole. The method does not estimate control delay for vehicles permitted to turn right on red, nor does it include these vehicles in the computation of capacity. It is recognized that vehicles making right turns on red (RTORs) increase intersection capacity. Thus, capacity, delay, and level of service may not be predicted accurately when RTORs occur. For the specific case in which a shared through and right-turn lane exists and RTORs are permitted, models were developed to estimate the number of RTORs that can be expected and the additional capacity that is realized. The models are deterministic and are consistent with the HCM methodology, are easily understood by analysts, produce reasonable results, and account for the probabilistic nature of RTORs occurring from shared lane approaches. The capacity model also demonstrates that, where RTORs are permitted and do occur, the HCM method results in an underestimation of capacity and an overestimation of delay for approaches containing shared lanes.

Recognition of Faces versus Greebles: A Case Study in Model Selection

An indicator tape is supported by idler pulleys in a flat position in front of a flat viewing window and is attached to a rotatable drum actuated by an amplifying linkage connected to a condition responsive means such as a Bourdon tube or diaphragm means for sensing a condition. The tape may be either endless or may have its other end attached to a spring-loaded return drum. The linkage connection between the bellows and the drum includes a zero adjustment and a range adjustment. A light transmitting plastic block is positioned behind the tape at the viewing window and is provided with cavities in which are mounted lights to illuminate the block and tape.