Promise and challenge of high-performance computing,with examples from molecular modelling

Computational modelling is one of the most significant developments in the practice of scientific inquiry in the 20th century. During the past decade, advances in computing technologies have increased the speed of computers by a factor of 100; an increase of a factor of 1000 can be expected in the next decade. These advances have, however, come at a price, namely, radical change(s) in computer architecture. Will computational scientists and engineers be able to harness the power offered by these high-performance computers to solve the most critical problems in science and engineering? In this paper, we discuss the challenges that must be addressed if we are to realize the benefits offered by high-performance computing. The task will not be easy; it will require revision or replacement of much of the software developed for vector supercomputers as well as advances in a number of key theoretical areas. Because of the pace of computing advances, these challenges must be met by close collaboration between computational scientists, computer scientists and applied mathematicians. The effectiveness of such a multidisciplinary approach is illustrated in a brief review of NWCHEM, a general-purpose computational chemistry code designed for parallel supercomputers.     

Study of the statistics of water vapor mixing ratio determined from Raman lidar measurements

The statistical properties of atmospheric water vapor mixing ratio (WVMR) determined as the ratio of Raman lidar signals backscattered from water vapor and nitrogen molecules are studied. It is shown that WVMR estimates can be biased by a small percentage at low signal photon-counting rates due to fluctuations in the nitrogen signal in the denominator of the ratio, the magnitude of the bias being linked to the signal-to-noise ratio of the nitrogen signal. This is particularly important when unbiased estimates are required as in the case of climate studies and global positioning system (GPS) signal calibration. Different bias corrections and a modified ratio formulation are proposed in order to correct or eliminate this bias. The method is successfully applied in processing signals obtained with an experimental Raman lidar system devoted to calibrate GPS signals for slant path delays. It is shown to reduce biases into negligible values in both WVMR and wet path delay estimates in the range interval of 0-7 km.     

Characterisation of high-performance cold bitumen emulsion mixtures for surface courses

Cold bitumen emulsion mixture (CBEM) is not yet widely used as a surface course around the world. In this study, 0/14-mm-size dense-graded surface course CBEMs have been investigated. The mechanical performance was evaluated in terms of stiffness modulus over 3 months and resistance to permanent deformation under three different stress levels (100, 200, 300 kPa), whilst durability evaluation was carried out in terms of resistance to moisture and frost damage. The study has also investigated the incorporation of low cement content (1%) with relatively sustainable by-product fillers, namely ground-granulated blast furnace slag (GGBS) and fly ash (FA) type 450-S on both mechanical and durability performance. A comparison has been carried out between the low and high cement content CBEM, as well as with respect to corresponding hot mix asphalt (HMA). The results revealed that the incorporation of GGBS and FA in CBEMs leads to superior performance, similar to CBEMs treated with high cement content and comparable to an equivalent HMA. Furthermore, GGBS replacement exhibited better performance than that of FA replacement. The findings suggest that the new sustainable types of CBEM can be developed for using as a surface layer for medium- to heavy-trafficked roads.     

Proteins and proteolysis in pre-term and term human milk and possible implications for infant formulae

Understanding the differences between the protein system of human milk and bovine milk is critical in the development of infant formulae. In this study, the proteins of bovine milk and a bovine-based whey-dominant infant formula were compared with those of human milk for infants born prematurely (pre-term) or at full term (term). The protein distribution of infant formula differed significantly from that of either type of human milk. A proteomic comparison between pre-term and term human milk showed a reduction of levels of β-casein and α_s-casein and appearance of additional products, corresponding to low molecular weight hydrolysis products of the caseins, in pre-term milk. Pre-term milk samples also had higher total nitrogen concentration and plasmin activity, consistent with the proteomic data. These results suggest the operation of a physiological mechanism that may adjust enzyme and/or protein expression to modify protein digestibility, and may facilitate design of infant formulae, closer to maternal milk, particularly for premature infants.     

Comparison of the principal proteins in bovine, caprine, buffalo, equine and camel milk

Proteomic analysis of bovine, caprine, buffalo, equine and camel milk highlighted significant interspecies differences. Camel milk was found to be devoid of β-lactoglobulin, whereas β-lactoglobulin was the major whey protein in bovine, buffalo, caprine, and equine milk. Five different isoforms of κ-casein were found in camel milk, analogous to the micro-heterogeneity observed for bovine κ-casein. Several spots observed in 2D-electrophoretograms of milk of all species could tentatively be identified as polypeptides arising from the enzymatic hydrolysis of caseins. The understanding gained from the proteomic comparison of these milks may be of relevance both in terms of identifying sources of hypoallergenic alternatives to bovine milk and detection of adulteration of milk samples and products.     

Genetic parameters for αS1-casein and αS2-casein phosphorylation isoforms in Dutch Holstein Friesian

Relative concentrations of α_S1-casein and α_S2-casein (α_S1-CN and α_S2-CN) phosphorylation isoforms vary considerably among milk of individual cows. We estimated heritabilities for α_S2-CN phosphorylation isoforms, determined by capillary zone electrophoresis from 1,857 morning milk samples, and genetic correlations among α_S2-CN phosphorylation isoforms in Dutch Holstein Friesian. To investigate if phosphorylation of α_S1-CN and α_S2-CN are due to the same genetic mechanism, we also estimated genetic correlations between α_S1-CN and α_S2-CN phosphorylation isoforms as well as the genetic correlations between the phosphorylation degrees (PD) of α_S1-CN and α_S2-CN defined as the proportion of isoforms with higher degrees of phosphorylation in total α_S1-CN and α_S2-CN, respectively. The intra-herd heritabilities for the relative concentrations of α_S2-CN phosphorylation isoforms were high and ranged from 0.54 for α_S2-CN-10P to 0.89 for α_S2-CN-12P. Furthermore, the high intra-herd heritabilities of α_S1-CN PD and α_S2-CN PD imply a strong genetic control of the phosphorylation process, which is independent of casein production. The genetic correlations between α_S2-CN phosphorylation isoforms are positive and moderate to high (0.33–0.90). Furthermore, the strong positive genetic correlation (0.94) between α_S1-CN PD and α_S2-CN PD suggests that the phosphorylation processes of α_S1-CN and α_S2-CN are related. This study shows the possibility of breeding for specific α_S1-CN and α_S2-CN phosphorylation isoforms, and relations between the phosphorylation degrees of α_S1-CN and α_S2-CN and technological properties of milk need to be further investigated to identify potential benefits for the dairy industry.     

HACCP plan proposal for a typical Brazilian peanut processing company

Peanuts are likely to be infested by fungi with consequent contamination by aflatoxin in post-harvest industries. A hazard analysis critical control point (HACCP) plan is proposed for a typical Brazilian post-harvest industry from raw in-shell reception to the unpeeled peanuts transportation. Codex Alimentarius Commission guidelines were followed, with four critical control points (CCP) for aflatoxin being identified. The process steps with highest probability of aflatoxin occurrence (risk) are the in-shell reception, the dried in-shell storage, and the unpeeled kernel storage. During the storage steps there is a lack of control of air moisture and temperature. Therefore, there is no option but to keep rigid monitoring and control over each CCP, and detour lots with high aflatoxin levels to either oil or seed production. Attempts to correlate the aflatoxin levels with the rainfall showed an irregular trend of the toxin level.     

Understanding statistical power in the context of applied research

Estimates of statistical power are widely used in applied research for purposes such as sample size calculations. This paper reviews the benefits of power and sample size estimation and considers several problems with the use of power calculations in applied research that result from misunderstandings or misapplications of statistical power. These problems include the use of retrospective power calculations and standardized measures of effect size. Methods of increasing the power of proposed research that do not involve merely increasing sample size (such as reduction in measurement error, increasing 'dose' of the independent variable and optimizing the design) are noted. It is concluded that applied researchers should consider a broader range of factors (other than sample size) that influence statistical power, and that the use of standardized measures of effect size should be avoided (except as intermediate stages in prospective power or sample size calculations).