microclass: an R-package for 16S taxonomy classification.

BACKGROUND: Taxonomic classification based on the 16S rRNA gene sequence is important for the profiling of microbial communities. In addition to giving the best possible accuracy, it is also important to quantify uncertainties in the classifications. RESULTS: We present an R package with tools for making such classifications, where the heavy computations are implemented in C++ but operated through the standard R interface. The user may train classifiers based on specialized data sets, but we also supply a ready-to-use function trained on a comprehensive training data set designed specifically for this purpose. This tool also includes some novel ways to quantify uncertainties in the classifications. CONCLUSIONS: Based on input sequences of varying length and quality, we demonstrate how the output from the classifications can be used to obtain high quality taxonomic assignments from 16S sequences within the R computing environment. The package is publicly available at the Comprehensive R Archive Network.

A retrospective analysis of cardiovascular outcomes in patients treated with ASV.

OBJECTIVES: The effect of long-term adaptive servo-ventilation (ASV) on cardiovascular mortality and admission rates in patients with chronic heart failure (CHF) and Cheyne-Stokes respiration (CSR) has not been much studied. The aim of this study was primarily to investigate whether ASV therapy significantly reduced these parameters. DESIGN: We included 75 CHF patients on optimal medication and CSR ≥25% of sleeping time, in New York Heart Association (NYHA) classes II-IV and left ventricular ejection fraction (LVEF) ≤ 45%. Thirty-one patients were treated with ASV for >3-18 months and 44 patients served as a control group. RESULTS: Seven deaths (16%) in the control group and one death (3%) in the ASV treatment group had cardiovascular etiology. There was no significant difference between the two groups regarding cardiovascular death (log rank p = 0.07; HR 0.18 (95% CI 0.02-1.44), p = 0.11) and combined cardiovascular death or readmissions, but there was a trend toward better outcome regarding cardiovascular event-free survival (log rank p = 0.06; HR 0.53 (95% CI 0.27-1.05). CONCLUSIONS: In CHF patients with CSR, 18 months ASV treatment did not significantly affect cardiovascular death or combined cardiovascular death or hospital admissions. But there was a trend toward better combined outcome.

micropan: an R-package for microbial pan-genomics.

BACKGROUND: A pan-genome is defined as the set of all unique gene families found in one or more strains of a prokaryotic species. Due to the extensive within-species diversity in the microbial world, the pan-genome is often many times larger than a single genome. Studies of pan-genomes have become popular due to the easy access to whole-genome sequence data for prokaryotes. A pan-genome study reveals species diversity and gene families that may be of special interest, e.g because of their role in bacterial survival or their ability to discriminate strains. RESULTS: We present an R package for the study of prokaryotic pan-genomes. The R computing environment harbors endless possibilities with respect to statistical analyses and graphics. External free software is used for the heavy computations involved, and the R package provides functions for building a computational pipeline. CONCLUSIONS: We demonstrate parts of the package on a data set for the gram positive bacterium Enterococcus faecalis. The package is free to download and install from The Comprehensive R Archive Network.

Comparing K-mer based methods for improved classification of 16S sequences.

BACKGROUND: The need for precise and stable taxonomic classification is highly relevant in modern microbiology. Parallel to the explosion in the amount of sequence data accessible, there has also been a shift in focus for classification methods. Previously, alignment-based methods were the most applicable tools. Now, methods based on counting K-mers by sliding windows are the most interesting classification approach with respect to both speed and accuracy. Here, we present a systematic comparison on five different K-mer based classification methods for the 16S rRNA gene. The methods differ from each other both in data usage and modelling strategies. We have based our study on the commonly known and well-used naïve Bayes classifier from the RDP project, and four other methods were implemented and tested on two different data sets, on full-length sequences as well as fragments of typical read-length. RESULTS: The difference in classification error obtained by the methods seemed to be small, but they were stable and for both data sets tested. The Preprocessed nearest-neighbour (PLSNN) method performed best for full-length 16S rRNA sequences, significantly better than the naïve Bayes RDP method. On fragmented sequences the naïve Bayes Multinomial method performed best, significantly better than all other methods. For both data sets explored, and on both full-length and fragmented sequences, all the five methods reached an error-plateau. CONCLUSIONS: We conclude that no K-mer based method is universally best for classifying both full-length sequences and fragments (reads). All methods approach an error plateau indicating improved training data is needed to improve classification from here. Classification errors occur most frequent for genera with few sequences present. For improving the taxonomy and testing new classification methods, the need for a better and more universal and robust training data set is crucial.

Corrigendum: Head and neck cancer treatment outcome prediction: a comparison between machine learning with conventional radiomics features and deep learning radiomics.

[This corrects the article DOI: 10.3389/fmed.2023.1217037.].

Deep learning with uncertainty estimation for automatic tumor segmentation in PET/CT of head and neck cancers: impact of model complexity, image processing and augmentation.

Objective.Target volumes for radiotherapy are usually contoured manually, which can be time-consuming and prone to inter- and intra-observer variability. Automatic contouring by convolutional neural networks (CNN) can be fast and consistent but may produce unrealistic contours or miss relevant structures. We evaluate approaches for increasing the quality and assessing the uncertainty of CNN-generated contours of head and neck cancers with PET/CT as input.Approach.Two patient cohorts with head and neck squamous cell carcinoma and baseline18F-fluorodeoxyglucose positron emission tomography and computed tomography images (FDG-PET/CT) were collected retrospectively from two centers. The union of manual contours of the gross primary tumor and involved nodes was used to train CNN models for generating automatic contours. The impact of image preprocessing, image augmentation, transfer learning and CNN complexity, architecture, and dimension (2D or 3D) on model performance and generalizability across centers was evaluated. A Monte Carlo dropout technique was used to quantify and visualize the uncertainty of the automatic contours.Main results. CNN models provided contours with good overlap with the manually contoured ground truth (median Dice Similarity Coefficient: 0.75-0.77), consistent with reported inter-observer variations and previous auto-contouring studies. Image augmentation and model dimension, rather than model complexity, architecture, or advanced image preprocessing, had the largest impact on model performance and cross-center generalizability. Transfer learning on a limited number of patients from a separate center increased model generalizability without decreasing model performance on the original training cohort. High model uncertainty was associated with false positive and false negative voxels as well as low Dice coefficients.Significance.High quality automatic contours can be obtained using deep learning architectures that are not overly complex. Uncertainty estimation of the predicted contours shows potential for highlighting regions of the contour requiring manual revision or flagging segmentations requiring manual inspection and intervention.

Non-linear shrinking of linear model errors.

BACKGROUND: Artificial neural networks (ANNs) can be a powerful tool for spectroscopic data analysis. Their ability to detect and model complex relations in the data may lead to outstanding predictive capabilities, but the predictions themselves are difficult to interpret due to the lack of understanding of the black box ANN models. ANNs and linear methods can be combined by first fitting a linear model to the data followed by a non-linear fitting of the linear model residuals using an ANN. This paper explores the use of residual modelling in high-dimensional data using modern neural network architectures. RESULTS: By combining linear- and ANN modelling, we demonstrate that it is possible to achieve both good model performance while retaining interpretations from the linear part of the model. The proposed residual modelling approach is evaluated on four high-dimensional datasets, representing two regression and two classification problems. Additionally, a demonstration of possible interpretation techniques are included for all datasets. The study concludes that if the modelling problem contains sufficiently complex data (i.e., non-linearities), the residual modelling can in fact improve the performance of a linear model and achieve similar performance as pure ANN models while retaining valuable interpretations for a large proportion of the variance accounted for. SIGNIFICANCE AND NOVELTY: The paper presents a residual modelling scheme using modern neural network architectures. Furthermore, two novel extensions of residual modelling for classification tasks are proposed. The study is seen as a step towards explainable AI, with the aim of making data modelling using artificial neural networks more transparent.

The effect of bio-electro-magnetic-energy-regulation therapy on sleep duration and sleep quality among elite players in Norwegian women's football.

The current study investigated if physical loads peak on game days and if Bio-Electro-Magnetic-Energy-Regulation (BEMER) therapy is affecting sleep duration and sleep quality on nights related to game nights among elite players in Norwegian women's elite football. The sample included 21 female football players from an elite top series club with a mean age of ~24 years (± 2.8). Sleep was measured every day over a period of 273 consecutive days with a Somnofy sleep monitor based on ultra-wideband (IR-UWB) pulse radar and Doppler technology. The current study was conducted as a quasi-experiment, where each player was their own control based on a control period that lasted for 3 months, and an experimental period that lasted for 5 months. Accordantly, the time each player spent with BEMER therapy was used as a control variable. Multivariate analyses of variance using FFMANOVA and univariate ANOVA with False Discovery Rate adjusted p-values show that physical performance (total distance, distance per minute, sprint meters >22.5 kmh, accelerations and decelerations) significantly peak on game day compared with ordinary training days and days related to game days. The results also show that sleep quantity and quality are significantly reduced on game night, which indicate disturbed sleep caused by the peak in physical load. Most sleep variables significantly increased in the experiment period, where BEMER therapy was used, compared to the control period before the introduction of BEMER therapy. Further, the analyses show that players who spent BEMER therapy >440 h had the most positive effects on their sleep, and that these effects were significantly compared to the players who used BEMER therapy <440 h. The findings are discussed based on the function of sleep and the different sleep stages have on recovery.

Head and neck cancer treatment outcome prediction: a comparison between machine learning with conventional radiomics features and deep learning radiomics.

Background: Radiomics can provide in-depth characterization of cancers for treatment outcome prediction. Conventional radiomics rely on extraction of image features within a pre-defined image region of interest (ROI) which are typically fed to a classification algorithm for prediction of a clinical endpoint. Deep learning radiomics allows for a simpler workflow where images can be used directly as input to a convolutional neural network (CNN) with or without a pre-defined ROI. Purpose: The purpose of this study was to evaluate (i) conventional radiomics and (ii) deep learning radiomics for predicting overall survival (OS) and disease-free survival (DFS) for patients with head and neck squamous cell carcinoma (HNSCC) using pre-treatment 18F-fluorodeoxuglucose positron emission tomography (FDG PET) and computed tomography (CT) images. Materials and methods: FDG PET/CT images and clinical data of patients with HNSCC treated with radio(chemo)therapy at Oslo University Hospital (OUS; n = 139) and Maastricht University Medical Center (MAASTRO; n = 99) were collected retrospectively. OUS data was used for model training and initial evaluation. MAASTRO data was used for external testing to assess cross-institutional generalizability. Models trained on clinical and/or conventional radiomics features, with or without feature selection, were compared to CNNs trained on PET/CT images without or with the gross tumor volume (GTV) included. Model performance was measured using accuracy, area under the receiver operating characteristic curve (AUC), Matthew's correlation coefficient (MCC), and the F1 score calculated for both classes separately. Results: CNNs trained directly on images achieved the highest performance on external data for both endpoints. Adding both clinical and radiomics features to these image-based models increased performance further. Conventional radiomics including clinical data could achieve competitive performance. However, feature selection on clinical and radiomics data lead to overfitting and poor cross-institutional generalizability. CNNs without tumor and node contours achieved close to on-par performance with CNNs including contours. Conclusion: High performance and cross-institutional generalizability can be achieved by combining clinical data, radiomics features and medical images together with deep learning models. However, deep learning models trained on images without contours can achieve competitive performance and could see potential use as an initial screening tool for high-risk patients.

Relating fatty acid composition in human fingertip blood to age, gender, nationality and n-3 supplementation in the Scandinavian population.

This study investigated data obtained from whole blood fatty acid (FA) composition of 3476 Norwegian and Swedish individuals, which provided background information including age, gender, nationality and self-motivated n-3 supplement consumption. The aim of this paper was to statistically relate this background information on the subjects to their whole blood FA profile, focusing mainly on the n-3 polyunsaturated FA (PUFA). Results showed that age had significant effects on the content of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) in blood lipids for the Norwegian individuals, while n-3 PUFA supplementation had a positive effect on EPA and DHA content in whole blood for the investigated population. Gender differences were also found for individual FA. A correlation also exists with previous studies on the FA profiling of blood lipids, further validating the test procedure.

Swiss knife partial least squares (SKPLS): One tool for modelling single block, multiblock, multiway, multiway multiblock including multi-responses and meta information under the ROSA framework.

In the domain of chemometrics and multivariate data analysis, partial least squares (PLS) modelling is a widely used technique. PLS gains its beauty by handling the high collinearity found in multivariate data by replacing highly covarying variables with common subspaces spanned by orthogonal latent variables. Furthermore, all can be achieved with simple steps of linear algebra requiring minimal computation power and time usage compared to current high-end computing and substantial hyperparameter tuning required by methods such as deep learning. PLS can be used for a wide variety of tasks, for example, single block modelling, multiblock modelling, multiway data modelling and for task such as regression and classification. Furthermore, new PLS based approaches can also incorporate meta information to improve the PLS subspace extraction. However, in the current scenario, there is a wide range of separate tools and codes available to perform different PLS tasks. Often when the user needs to perform a new PLS task, they need to start with a separate mathematical implementation of the PLS techniques. This study aims to provide a single solution, i.e., the Swiss knife PLS (SKPLS) modelling approach to enable a single mathematical implementation to perform analyses of single block, multiblock, multiway, multiblock multiway, multi-response, and incorporation of meta information in PLS modelling. It contains all that is needed for any PLS practitioner to perform both classification and regression tasks. The SKPLS backbone is the stepwise PLS strategy called response oriented sequential alternation (ROSA) which we generalize to enable all the mentioned analysis possibilities. The basic structure of the algorithm is highlighted, and some example cases of performing single block, multiblock, multiway, multiblock multiway, multi-response PLS modelling and the incorporation of meta information in PLS modelling are included.

META-PLS modelling: An integrated approach to automatic model optimization for near-infrared spectra.

Predictive latent space near-infrared (NIR) spectral modelling with PLS (Partial Least Squares) has two main tasks that require user input to achieve optimal models. The first is the selection of the optimal pre-processing of NIR spectra and the second is the selection of the optimal number of PLS model components assuming the data is outlier free. Often the two tasks are performed in an exhaustive search to find the best pre-processing as well as the optimal number of model components. We propose a novel approach called meta partial least square (META-PLS) which drops the need for both the pre-processing optimisation and exhaustive search for optimal model components. We utilise the stepwise nature of the PLS algorithm to learn complementary information from different pre-processed forms of the same data set as performed in multiblock pre-processing ensemble models to avoid pre-processing selection but receive help from the pre-processing ensembles, and deploy a weighted randomisation test to decide the optimal number of model components automatically. The performance of the approach for performing automatic NIR spectral modelling is demonstrated with several real data sets.

Encoder-decoder neural networks for predicting future FTIR spectra - application to enzymatic protein hydrolysis.

In the process of converting food-processing by-products to value-added ingredients, fine grained control of the raw materials, enzymes and process conditions ensures the best possible yield and economic return. However, when raw material batches lack good characterization and contain high batch variation, online or at-line monitoring of the enzymatic reactions would be beneficial. We investigate the potential of deep neural networks in predicting the future state of enzymatic hydrolysis as described by Fourier-transform infrared spectra of the hydrolysates. Combined with predictions of average molecular weight, this provides a flexible and transparent tool for process monitoring and control, enabling proactive adaption of process parameters.

Deep learning-enabled Inference of 3D molecular absorption distribution of biological cells from IR spectra.

Infrared spectroscopy delivers abundant information about the chemical composition, as well as the structural and optical properties of intact samples in a non-destructive manner. We present a deep convolutional neural network which exploits all of this information and solves full-wave inverse scattering problems and thereby obtains the 3D optical, structural and chemical properties from infrared spectroscopic measurements of intact micro-samples. The proposed model encodes scatter-distorted infrared spectra and infers the distribution of the complex refractive index function of concentrically spherical samples, such as many biological cells. The approach delivers simultaneously the molecular absorption, sample morphology and effective refractive index in both the cell wall and interior from a single measured spectrum. The model is trained on simulated scatter-distorted spectra, where absorption in the distinct layers is simulated and the scatter-distorted spectra are estimated by analytic solutions of Maxwell's equations for samples of different sizes. This allows for essentially real-time deep learning-enabled infrared diffraction micro-tomography, for a large subset of biological cells.

Oxidative Stability of Polyunsaturated Edible Oils Mixed With Microcrystalline Cellulose.

The oxidative stability of mixtures of edible oils containing polyunsaturated fatty acids (PUFA) and microcrystalline cellulose (MCC) was investigated. The mixtures studied consisted of oils of either camelina (CAM), cod liver (CLO), or salmon (SO) mixed with either colloidal or powdered MCC. A 50:50 (w/w) ratio of oil:MCC resulted in an applicable mixture containing high levels of PUFA edible oil and dietary fiber. The oxidative stability of the formulated mixtures and the pure oils was investigated over a period of 28 days. The peroxide value (PV) was assessed as a parameter for primary oxidation products and dynamic headspace gas chromatography mass spectrometry (GC/MS) was used to analyze secondary volatile organic compounds (VOC). CAM and the respective mixtures were oxidatively stable at both 4 and 22 °C during the storage period. The marine oils and the respective mixtures were stable at 4 °C. At 22 °C, an increase in hydroperoxides was found, but no increase in VOC was detected during the time-frame investigated. At 42 °C, prominent increases in PV and VOC were found for all oils and mixtures. Hexanal, a common marker for the degradation of n-6 fatty acids, propanal and 2,4-heptadienal (E,E), common indicators for the degradation of n-3 fatty acids, were among the volatiles detected in the headspace of oils and mixtures. This study showed that a mixture containing a 50:50 ratio of oil:MCC can be obtained by a low-tech procedure that does not induce oxidation when stored at low temperatures during a period of 1 month.

Effect of Liquid Absorbent Pads and Packaging Parameters on Drip Loss and Quality of Chicken Breast Fillets.

Visible liquid inside food packages is perceived as unattractive to consumers, and may result in food waste-a significant factor that can compromise sustainability in food value chains. However, an absorber with overdimensioned capacity may cause alterations in texture and a dryer product, which in turn may affect consumers' satisfaction and repurchase. In this study we compared the effect of a number of liquid absorbent pads in combination with headspace gas composition (60% CO2/40% N2 and 75% O2/25% CO2) and gas-to-product volume ratio (g/p) on drip loss and quality of fresh chicken breast fillets. A significant increase in drip loss with an increasing number of liquid absorbent pads was documented. The increase was more pronounced in 60% CO2/40% N2 compared to 75% O2/25% CO2. By comparing packaging variants with a different number of liquid absorbent pads, a higher drip loss for all tested was found at g/p 1.8 compared to g/p 2.9. Total viable counts (TVC) were independent of whether there was free liquid in contact with the product, and TVC was independent of gas composition. Differentiation between the gas compositions was seen for specific bacterial analyses. While significant changes were observed using texture analysis, sensory evaluation of the chicken breast fillets did not show any negative effect in texture related attributes. This study demonstrates the importance of optimized control of meat drip loss, as product-adjusted liquid absorption may affect economy, food quality, and consumer satisfaction, as well as food waste.

Cerebrospinal fluid proteome shows disrupted neuronal development in multiple sclerosis.

Despite intensive research, the aetiology of multiple sclerosis (MS) remains unknown. Cerebrospinal fluid proteomics has the potential to reveal mechanisms of MS pathogenesis, but analyses must account for disease heterogeneity. We previously reported explorative multivariate analysis by hierarchical clustering of proteomics data of MS patients and controls, which resulted in two groups of individuals. Grouping reflected increased levels of intrathecal inflammatory response proteins and decreased levels of proteins involved in neural development in one group relative to the other group. MS patients and controls were present in both groups. Here we reanalysed these data and we also reanalysed data from an independent cohort of patients diagnosed with clinically isolated syndrome (CIS), who have symptoms of MS without evidence of dissemination in space and/or time. Some, but not all, CIS patients had intrathecal inflammation. The analyses reported here identified a common protein signature of MS/CIS that was not linked to elevated intrathecal inflammation. The signature included low levels of complement proteins, semaphorin-7A, reelin, neural cell adhesion molecules, inter-alpha-trypsin inhibitor heavy chain H2, transforming growth factor beta 1, follistatin-related protein 1, malate dehydrogenase 1 cytoplasmic, plasma retinol-binding protein, biotinidase, and transferrin, all known to play roles in neural development. Low levels of these proteins suggest that MS/CIS patients suffer from abnormally low oxidative capacity that results in disrupted neural development from an early stage of the disease.

Obstructive sleep apnea versus central sleep apnea: prognosis in systolic heart failure.

BACKGROUND: In chronic heart failure (CHF), obstructive sleep apnea (OSA) and Cheyne-Stokes respiration (CSR) are associated with increased mortality. The present study aimed to evaluate the prognostic effect of CSR compared to OSA, in otherwise similar groups of CHF patients. METHODS: Screening for sleep-disordered breathing (SDB) was conducted among patients with CHF of New York Heart Association (NYHA) class II-IV, and left ventricular ejection fraction (LVEF) of ≤45%. The study included 43 patients (4 women) with >25% CSR during sleeping time, and 19 patients (2 women) with OSA and an apnea-hypopnea index (AHI) of ≥6. Patients were followed for a median of 1,371 days. The primary endpoint was mortality, and the secondary endpoint was combined mortality and hospital admissions. RESULTS: Baseline parameters did not significantly differ between groups, but CSR patients were older and had higher AHI values than OSA patients. Five OSA patients (26%) died, and 14 (74%) met the combined end-point of death or hospitalization. CSR patients had significantly higher risk for both end-points, with 23 (53%) deaths [log-rank P=0.040; HR, 2.70 (1.01-7.22); P=0.047] and 40 (93%) deaths or readmissions [log-rank P=0.029; HR, 1.96 (1.06-3.63); P=0.032]. After adjustment for confounding risk factors, the association between CSR and death remained significant [HR, 4.73 (1.10-20.28); P=0.037], hospital admission rates were not significantly different. CONCLUSIONS: Among patients with CHF, CSR was associated with higher mortality than OSA independently of age and cardiac systolic function. CSR was also an age-independent predictor of unfavorable outcome, but hospital admission rates were not significantly different between the two groups after adjustment.

SO-PLS as an alternative approach for handling multi-dimensionality in modelling different aspects of consumer expectations.

In the development of sensory and consumer science, data are often collected in several blocks responding to different aspects of consumer experience. Sometimes the task of organizing the data and explaining their relation is non-trivial, especially when considering structural (casual) relationship between data sets. In this sense, PLS path modelling (PLS-PM) has been found as a good tool to model such relations, but this approach faces some issues regarding the assumption of uni-dimensionality of consumers' data blocks. Sequential Orthogonalised PLS path modelling (SO-PLS-PM) has been proposed as an alternative approach to handle the multi-dimensionality and to explain the relations between the original data blocks without any preprocessing of the data. This study aims at comparing the efficacy of SO-PLS-PM and PLS-PM (together with splitting blocks into uni-dimensional sub-blocks) for handling multi-dimensionality. Data sets from two satiety perception studies (yoghurt, biscuit) have been used as illustrations. The main novelty of this paper lies in underlining and solving a major, but little studied problem, related to the assumption of one-dimensional blocks in PLS-PM. The findings from the comparisons indicated that the two approaches (PLS-PM and SO-PLS-PM) highlighted the same main trends for the less complex samples (yoghurt samples): liking was the essential driver of satiation perception and portion size selection; while satiation mainly predicted satiety perception. For the more complex data set - from a sensory perspective - (biscuit samples), the relations between data blocks in PLS-PM model was difficult to interpret, whereas they were well explained by SO-PLS-PM. This underlines the ability of SO-PLS-PM to model multi-dimensional data sets without requiring any preprocessing steps.

Secondary Lactic Acid Bacteria Fermentation with Wood-Derived Xylooligosaccharides as a Tool To Expedite Sour Beer Production.

Xylooligosaccharides (XOS) from woody biomass were evaluated as a substrate for secondary lactic acid bacteria (LAB) fermentation in sour beer production. XOS were extracted from birch (Betula pubescens) and added to beer to promote the growth of Lactobacillus brevis BSO 464. Growth, pH, XOS degradation, and metabolic products were monitored throughout fermentations, and the final beer was evaluated sensorically. XOS were utilized, metabolic compounds were produced (1800 mg/L lactic acid), and pH was reduced from 4.1 to 3.6. Secondary fermentation changed sensory properties significantly, and the resulting sour beer was assessed as similar to a commercial reference in multiple attributes, including acidic taste. Overall, secondary LAB fermentation induced by wood-derived XOS provided a new approach to successfully produce sour beer with reduced fermentation time (from 1-3 years to 4 weeks). The presented results demonstrate how hemicellulosic biomass can be valorized for beverage production and to obtain sour beer with improved process control.