Parallel neural net training on the KSR1

In modern day pattern recognition, neural nets are used extensively. General use of a feedforward neural net consists of a training phase followed by a classification phase. Classification of an unknown test vector is very fast and only consists of the propagation of the test vector through the neural net. Training involves an optimization procedure and is very time-consuming since a feasible local minimum is sought in high-dimensional weight space. In this paper we present an analysis of a parallel implementation of the backpropagation training algorithm using conjugate gradient optimization for a three-layered, feedforward neural network, on the KSR1 parallel shared-memory machine. We implement two parallel neural net training versions on the KSR1, one using native code, the other using P4, a library of macros and functions. A speedup model is presented which we use to clarify our experimental results. We identify the general requirements which render the parallel implementation useful, compared to the sequential execution of the same neural net training procedure. We determine the usefulness of a library of functions (such as P4) developed to ease the task of the programmer. Using experimental results we further identify the limits in processor utilization for our parallel training algorithm.     

A Customer-to-carrier Gateway for Multiple Services

This paper describes a customer-to-carrier prototype, based on a protocol-independent interface specification and the OMNIPoint Network Management Forum service management concept.     

Genome-wide bacterial toxicity screening uncovers the mechanisms of toxicity of a cationic polystyrene nanomaterial

By exploiting a genome-wide collection of bacterial single-gene deletion mutants, we have studied the toxicological pathways of a 60-nm cationic (amino-functionalized) polystyrene nanomaterial (PS-NH(2)) in bacterial cells. The IC(50) of commercially available 60 nm PS-NH(2) was determined to be 158 μg/mL, the IC(5) is 108 μg/mL, and the IC(90) is 190 μg/mL for the parent E. coli strain of the gene deletion library. Over 4000 single nonessential gene deletion mutants of Escherichia coli were screened for the growth phenotype of each strain in the presence and absence of PS-NH(2). This revealed that genes clusters in the lipopolysaccharide biosynthetic pathway, outer membrane transport channels, ubiquinone biosynthetic pathways, flagellar movement, and DNA repair systems are all important to how this organism responds to cationic nanomaterials. These results, coupled with those from confirmatory assays described herein, suggest that the primary mechanisms of toxicity of the 60-nm PS-NH(2) nanomaterial in E. coli are destabilization of the outer membrane and production of reactive oxygen species. The methodology reported herein should prove generally useful for identifying pathways that are involved in how cells respond to a broad range of nanomaterials and for determining the mechanisms of cellular toxicity of different types of nanomaterials.     

Monitoring Insulin Aggregation via Capillary Electrophoresis

Early stages of insulin aggregation, which involve the transient formation of oligomeric aggregates, are an important aspect in the progression of Type II diabetes and in the quality control of pharmaceutical insulin production. This study is the first to utilize capillary electrophoresis (CE) with ultraviolet (UV) detection to monitor insulin oligomer formation at pH 8.0 and physiological ionic strength. The lag time to formation of the first detected species in the aggregation process was evaluated by UV-CE and thioflavin T (ThT) binding for salt concentrations from 100 mM to 250 mM. UV-CE had a significantly shorter (5-8 h) lag time than ThT binding (15-19 h). In addition, the lag time to detection of the first aggregated species via UV-CE was unaffected by salt concentration, while a trend toward an increased lag time with increased salt concentration was observed with ThT binding. This result indicates that solution ionic strength impacts early stages of aggregation and β-sheet aggregate formation differently. To observe whether CE may be applied for the analysis of biological samples containing low insulin concentrations, the limit of detection using UV and laser induced fluorescence (LIF) detection modes was determined. The limit of detection using LIF-CE, 48.4 pM, was lower than the physiological insulin concentration, verifying the utility of this technique for monitoring biological samples. LIF-CE was subsequently used to analyze the time course for fluorescein isothiocyanate (FITC)-labeled insulin oligomer formation. This study is the first to report that the FITC label prevented incorporation of insulin into oligomers, cautioning against the use of this fluorescent label as a tag for following early stages of insulin aggregation.     

Chemistry of agaroids,carrageenans and furcellarans

The differences in the fine structure of these algal polysaccharides which are all polydisperse linear galactans are described. The effect of these differences on the conformation of the macromolecules and consequently on the physical properties, in particular the gelling properties, of the individual polymers is discussed. The evidence that gel formation occurs through the association of chain segments into a three dimensional framework which is held together by hydrogen bonds is reviewed.     

Investigating different factor structures of the Psychopathy Checklist: Youth version: Confirmatory factor analytic findings.

There has been a recent push to extend the construct of psychopathy into adolescence, primarily as a result of the impressive reliability, validity, and utility of this construct in samples of adults. The value of this work rests, however, on creating an equally reliable and valid assessment tool for adolescents. One promising measure is the Psychopathy Checklist: Youth Version (A. E. Forth, D. S. Kosson, & R. D. Hare, 2003). The current study uses a large, diverse sample of serious adolescent offenders to assess the overall fit of various underlying factor structures of this measure and to test the equivalence of these models across sex and race/ethnicity. The results suggest that either a 3- or 4-factor model provides the best overall fit and that these models are invariant across sex and race/ethnicity. The decision to use the 3- or 4-factor model will likely hinge on researchers' underlying conceptualization of psychopathy, specifically whether antisocial behavior is viewed as a core feature of this construct. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Cleaning UF membranes with simple and formulated solutions

The ultrafiltration membranes fouled by proteins are typically cleaned by consecutive soaking in alkali, surfactant and oxidizing solutions. We combined all three chemicals into a formulated cleaning agent and examined its efficiency to restore the water flux without damaging the membrane or enhancing protein fouling. Lower concentrations of ingredients in the composite were required to restore the water flux to the initial level. The membrane after the cleaning remained intact and subsequent filtration left trace protein amount on the membrane surface. We propose a mechanism of cleaning based on a fast penetration of alkali and oxidizing ingredients through a fouling layer and efficient micellation of detached foulants with surfactant.     

Molecular mechanisms of epigenetic variation in plants

Natural variation is defined as the phenotypic variation caused by spontaneous mutations. In general, mutations are associated with changes of nucleotide sequence, and many mutations in genes that can cause changes in plant development have been identified. Epigenetic change, which does not involve alteration to the nucleotide sequence, can also cause changes in gene activity by changing the structure of chromatin through DNA methylation or histone modifications. Now there is evidence based on induced or spontaneous mutants that epigenetic changes can cause altering plant phenotypes. Epigenetic changes have occurred frequently in plants, and some are heritable or metastable causing variation in epigenetic status within or between species. Therefore, heritable epigenetic variation as well as genetic variation has the potential to drive natural variation.     

First year growth of preterm infants fed standard compared to marine oil n−3 supplemented formula

Very low birth weight (VLBW) infants (748–1390 g, n=65) were randomly assigned to receive control or marine oil-supplemented formula when they achieved intakes >454 kJ (110 kcal)/kg/d of a formula designed for VLBW infants. Study formulas with or without marine oil were provided until 79 wk of postconceptional age (PCA), first in a formula designed for preterm infants followed by a formula designed for term infants. Infants were studied at regular intervals through 92 wk PCA. Weight, length, and head circumference were determined by standardized prodedures and normalized to the National Center for Health Statistics figures for growth of infants born at term of the same age and gender. Mean normalized weight, weight-to-length, and head circumference were greatest at 48 wk and decreased thereafter. The decline in normalized weight was greater in infants fed the marine oil-supplemented formula. Beginning at 40 wk, marine oil-supplemented infants compared to controls had significantly poorer Z-scores for weight, length and head circumference. In addition, birth order (negatively) and maternal height (positively) influenced weight and length achievement in infancy as shown previously in infants born at term. Based on a paper presented at the Symposium on Milk Lipids held at the AOCS Annual Meeting, Baltimore, MD, April 1990.     

Dietary intake of essential and long-chain polyunsaturated fatty acids in pregnancy

The dietary intake of EFA and long-chain PUFA (LCPUFA) by women with (n=14) and without (n=31) gestational diabetes mellitus (GDM) was determined by repeated 24-h recalls. Women with GDM consumed significantly more energy as fat compared with women who had uncomplicated pregnancies; absolute dietary fat did not differ. Dietary n−3 LCPUFA was substantially lower than the current recommendation for pregnancy, whereas intake of saturated FA (SFA) exceeded it. We conclude that replacing dietary sources of SFA with those of EFA and LCPUFA, especially n−3 LCPUFA, would benefit the dietary fat profiles of all pregnant women.