Evaluating the performance of machine learning methods and variable selection methods for predicting difficult-to-measure traits in Holstein dairy cattle using milk infrared spectral data

Fourier-transform infrared (FTIR) spectroscopy is a powerful high-throughput phenotyping tool for predicting traits that are expensive and difficult to measure in dairy cattle. Calibration equations are often developed using standard methods, such as partial least squares (PLS) regression. Methods that employ penalization, rank-reduction, and variable selection, as well as being able to model the nonlinear relations between phenotype and FTIR, might offer improvements in predictive ability and model robustness. This study aimed to compare the predictive ability of 2 machine learning methods, namely random forest (RF) and gradient boosting machine (GBM), and penalized regression against PLS regression for predicting 3 phenotypes differing in terms of biological meaning and relationships with milk composition (i.e., phenotypes measurable directly and not directly in milk, reflecting different biological processes which can be captured using milk spectra) in Holstein-Friesian cattle under 2 cross-validation scenarios. The data set comprised phenotypic information from 471 Holstein-Friesian cows, and 3 target phenotypes were evaluated: (1) body condition score (BCS), (2) blood β-hydroxybutyrate (BHB, mmol/L), and (3) κ-casein expressed as a percentage of nitrogen (κ-CN, % N). The data set was split considering 2 cross-validation scenarios: samples-out random in which the population was randomly split into 10-folds (8-folds for training and 1-fold for validation and testing); and herd/date-out in which the population was randomly assigned to training (70% herd), validation (10%), and testing (20% herd) based on the herd and date in which the samples were collected. The random grid search was performed using the training subset for the hyperparameter optimization and the validation set was used for the generalization of prediction error. The trained model was then used to assess the final prediction in the testing subset. The grid search for penalized regression evidenced that the elastic net (EN) was the best regularization with increase in predictive ability of 5%. The performance of PLS (standard model) was compared against 2 machine learning techniques and penalized regression using 2 cross-validation scenarios. Machine learning methods showed a greater predictive ability for BCS (0.63 for GBM and 0.61 for RF), BHB (0.80 for GBM and 0.79 for RF), and κ-CN (0.81 for GBM and 0.80 for RF) in samples-out cross-validation. Considering a herd/date-out cross-validation these values were 0.58 (GBM and RF) for BCS, 0.73 (GBM and RF) for BHB, and 0.77 (GBM and RF) for κ-CN. The GBM model tended to outperform other methods in predictive ability around 4%, 1%, and 7% for EN, RF, and PLS, respectively. The prediction accuracies of the GBM and RF models were similar, and differed statistically from the PLS model in samples-out random cross-validation. Although, machine learning techniques outperformed PLS in herd/date-out cross-validation, no significant differences were observed in terms of predictive ability due to the large standard deviation observed for predictions. Overall, GBM achieved the highest accuracy of FTIR-based prediction of the different phenotypic traits across the cross-validation scenarios. These results indicate that GBM is a promising method for obtaining more accurate FTIR-based predictions for different phenotypes in dairy cattle.     

Effects of nisin concentration on properties of gelatin film-forming solutions and their films

The addition of nisin into a gelatin matrix can change properties of the film. The aim of this work was to develop gelatin-based films containing different nisin concentrations in order to study their influence on the film's antimicrobial and physical properties and their rheological properties as a film-forming solution (FFS). The FFS was characterised by rheological assays, and the gelatin-based active films were characterised and assessed by the effects of nisin concentrations on their various properties, including antimicrobial activity. Nisin's concentration affected not only its viscoelastic properties of FFS but also its film solubility in water, film surface roughness and light barrier. The addition of nisin also slightly modified the water contact angle and the mechanical properties of the gelatin films. Finally, the films demonstrated activity against Staphylococcus aureus and Listeria monocytogenes at concentrations above 56 mg of nisin g⁻¹ of gelatin.     

Platelets: Still a Therapeutical Target for Haemostatic Disorders

Platelets are cytoplasmatic fragments from bone marrow megakaryocytes present in blood. In this work, we review the basis of platelet mechanisms, their participation in syndromes and in arterial thrombosis, and their potential as a target for designing new antithrombotic agents. The option of new biotechnological sources is also explored.     

Volume-Independent Sodium Toxicity in Peritoneal Dialysis: New Insights from Bench to Bed

Sodium overload is common in end-stage kidney disease (ESKD) and is associated with increased cardiovascular mortality that is traditionally considered a result of extracellular volume expansion. Recently, sodium storage was detected by Na23 magnetic resonance imaging in the interstitial tissue of the skin and other tissues. This amount of sodium is osmotically active, regulated by immune cells and the lymphatic system, escapes renal control, and, more importantly, is associated with salt-sensitive hypertension. In chronic kidney disease, the interstitial sodium storage increases as the glomerular filtration rate declines and is related to cardiovascular damage, regardless of the fluid overload. This sodium accumulation in the interstitial tissues becomes more significant in ESKD, especially in older and African American patients. The possible negative effects of interstitial sodium are still under study, though a higher sodium intake might induce abnormal structural and functional changes in the peritoneal wall. Interestingly, sodium stored in the interstial tissue is not unmodifiable, since it is removable by dialysis. Nevertheless, the sodium removal by peritoneal dialysis (PD) remains challenging, and new PD solutions are desirable. In this narrative review, we carried out an update on the pathophysiological mechanisms of volume-independent sodium toxicity and possible future strategies to improve sodium removal by PD.     

MASS TRANSFER AND DIFFUSION COEFFICIENT DETERMINATION IN THE WET AND DRY SALTING OF MEAT

ABSTRACT

Dehydrated salted meat is widely used in Brazil as a very important source of animal protein. The main objective of this kind of processing is water removal. initially by osmotic pressure changes and then by drying, resulting in a product with intermediate moisture levels.

In this work, mass transfer and salt diffusion in pieces of meat submitted to wet and dry salting were studied. Slabs of beef m. trapezius with an infinite plate geometry were salted in a NaCl saturated solution or in a dry salt bed, at two temperatures (10 and 20°C) and different time exposures (120 min and 96 hours). Equilibration studies were extended up to six days.

It was observed that water loss increased with salt uptake, for increasing periods of times. At 20°C the moisture loss was higher than it was at 10°C in both salting processes. On the other hand, the kinetics of salt uptake and moisture loss were of greater importance in the process of dry salting than in that of wet salting.

The salt diffusion coefficient for wet salting was 0.26 × 10^?10m²/s at20°C and 0.25 × 10^?10 m²/s at 10°C and for the dry salting the values were 19.37 × 10^?10 m²/s at 20°C and 17.21 × 10^?10 m²/s at 10°C.     

Homonuclear Decoupling of 1H Dipolar Interactions in Solids by means of Heteronuclear Recoupling

Line narrowing has been traditionally achieved in solid-state ¹H NMR spectroscopy by applying pulse sequences that combine multiple-pulse operations with magic-angle spinning (MAS), to effectively average out the dipole dipole homonuclear Hamiltonian. The present study explores a new alternative that departs from the usual concept of directly acting on the strongly coupled spins with radiofrequency pulses; instead, we seek to achieve a net homonuclear dipolar decoupling in solids by exploring the reintroduction of MAS-averaged heteronuclear dipolar couplings between the ¹H nuclei and directly bonded ¹³C or ¹⁵N nuclei. This recoupling anti-recoupling (RaR) scheme thus relies on the recoupling of the dipolar interaction with heteronuclear spins, which, under fast MAS, will exceed the strength and will not commute with the homonuclear ¹H ¹H coupling one is intending to average out. Subsequent removal (“antiRecoupling”) of these heteronuclear interactions can lead to narrowed ¹H resonances, without ever pulsing on the aforementioned channel. The line-narrowing properties of RaR are illustrated with numerical simulations and with experiments on model organic solids.     

Liquid limit of low- to medium-plasticity soils: comparison between Casagrande cup and cone penetrometer test

For liquid limit determination, the European Standards require the cone penetrometer technique or, alternatively, the Casagrande cup method; but the latter is widely used in practice. As most European soil classification systems are not adapted to the new European Standards, a systematic study was carried out to verify the differences between the two techniques for liquid limit determination, examining clayey soils from various geologic units. Comparison of 100 liquid limit values ranging between 20 and 50% indicates that the liquid limit obtained by the fall cone method is generally 2.2% points higher than that obtained by the hard base Casagrande apparatus. Comparison of data from different sources consistently indicates that the equation proposed in this study may be useful in revising existing soil classification systems, avoiding problems of soil classification due to the technique used for liquid limit determination.     

A theoretical model of the relationship between the <Emphasis Type=Italic>h</Emphasis>-index and other simple citation indicators

In this study, we quantitatively compared the impact of mission-oriented research grants and curiosity-driven grants on the diversity of research subjects in Japan. First, we examined data for Japanese principal investigators receiving research funding between 2000 and 2010 in the field of nanotechnology and materials science, and identified groups of researchers whose publication performance was positively affected by the mission-oriented grant, CREST. We then compared the effect of CREST with that of the curiosity-driven grant, KAKENHI. The analysis uses both propensity score matching and difference in differences (PSM-DID) methodologies. Our results show that for participants in the CREST program there was an increase in number of publications of more than 10% per year, for periods of both 5 and 3 years after the funding ended, even though the observed average effect on citation was not statistically significant. Second, we evaluated the diversity of research subjects through analysis of the distribution of the classification codes applied to articles published between 1996 and 2013, utilizing the J-Global database, which has the finest granularity of category among existing bibliographic scientific publication databases. Research subjects were better conserved under the mission-oriented program than the curiosity-driven one, a finding contrary to predictions of conventional theory. We also found that under mission-oriented funding, there was an increase in diversity in the sense of marginal utility. These findings should be of use in the “diversity-aware” design of programs for the funding of fundamental research.