Search

The transition from wakefulness to sleep occurs when the core body temperature decreases. The latter is facilitated by an increase in the cutaneous blood flow, which dissipates internal heat into the micro-environment surrounding the sleeper's body. The rise in cutaneous blood flow near sleep onset causes the distal (hands and feet) and proximal (abdomen) temperatures to increase by about 1 °C and 0.5 °C, respectively. Characterizing the dynamics of skin temperature changes throughout sleep phases and understanding its relationship with sleep quality requires a means to unobtrusively and longitudinally estimate the skin temperature. Leveraging the data from a temperature sensor strip (TSS) with five individual temperature sensors embedded near the surface of a smart bed's mattress, we have developed an algorithm to estimate the distal skin temperature with a minute-long temporal resolution. The data from 18 participants who recorded TSS and ground-truth temperature data from sleep during 14 nights at home and 2 nights in a lab were used to develop an algorithm that uses a two-stage regression model (gradient boosted tree followed by a random forest) to estimate the distal skin temperature. A five-fold cross-validation procedure was applied to train and validate the model such that the data from a participant could only be either in the training or validation set but not in both. The algorithm verification was performed with the in-lab data. The algorithm presented in this research can estimate the distal skin temperature at a minute-level resolution, with accuracy characterized by the mean limits of agreement [-0.79 to +0.79 °C] and mean coefficient of determination R2=0.87. This method may enable the unobtrusive, longitudinal and ecologically valid collection of distal skin temperature values during sleep. Therelatively small sample size motivates the need for further validation efforts.

The ability to estimate the probability of a drug to receive approval in clinical trials provides natural advantages to optimizing pharmaceutical research workflows. Success rates of clinical trials have deep implications for costs, duration of development, and under pressure due to stringent regulatory approval processes. We propose a machine learning approach that can predict the outcome of the trial with reliable accuracies, using biological activities, physicochemical properties of the compounds, target-related features, and NLP-based compound representation. In the above list, biological activities have never been used as an independent variable towards the prediction of clinical trial outcomes. We have extracted the drug-disease pair from clinical trials and mapped target(s) to that pair using multiple data sources. Empirical results demonstrate that ensemble learning outperforms independently trained, small-data ML models. We report results and inferences derived from a Random forest classifier with an average accuracy of 93%, and an F1 score of 0.96 for the "Pass" class. "Pass" refers to one of the two classes (Pass/Fail) of all clinical trials, and the model performed well in predicting the "Pass" category. Through the analysis of feature contributions to predictive capability, we have demonstrated that bioactivity plays a statistically significant role in predicting clinical trial outcome. A significant effort has gone into the production of the dataset that, for the first time, integrates clinical trial information with protein targets. Cleaned, organized, integrated data and code to map these entities, created as a part of this work, are available open-source. This reproducibility and the freely available code ensure that researchers with access to deep curated and proprietary clinical trial databases (we only use open-source data in this study) can further expand the scope of the results., (© 2022 John Wiley & Sons Ltd.)

Motivation: The standard approach to identifying peptides based on accurate mass and elution time (AMT) compares profiles obtained from a high resolution mass spectrometer to a database of peptides previously identified from tandem mass spectrometry (MS/MS) studies. It would be advantageous, with respect to both accuracy and cost, to only search for those peptides that are detectable by MS (proteotypic)., Results: We present a support vector machine (SVM) model that uses a simple descriptor space based on 35 properties of amino acid content, charge, hydrophilicity and polarity for the quantitative prediction of proteotypic peptides. Using three independently derived AMT databases (Shewanella oneidensis, Salmonella typhimurium, Yersinia pestis) for training and validation within and across species, the SVM resulted in an average accuracy measure of 0.8 with a SD of <0.025. Furthermore, we demonstrate that these results are achievable with a small set of 12 variables and can achieve high proteome coverage., Availability: http://omics.pnl.gov/software/STEPP.php., Supplementary Information: Supplementary data are available at Bioinformatics online.

Scalability is an important desirable property of video codecs. Wavelet-based motion-compensated temporal filtering provides the most powerful scheme for scalable video coding and provides high-compression efficiency that competes with the current state of art codecs. However, the delay introduced by the temporal filtering schemes is sometimes very high, which makes them unsuitable for many real-time applications. In this paper, ue propose a new temporal filter set to minimize delay in 3-D wavelet-based video coding. The new filter set gives a performance at par with existing longer filters. The length of the filter can vary from two to any number of frames depending on delay requirements. If the frames are processed as separate groups of frames (GOFs), the proposed filter set will not have any boundary effects at the GOF. Experimental results are presented and conclusions are drawn.

In this paper, we present a different process for the optimal solution to mixed constraints transportation problem with imprecise coefficients without changing to its crisp equivalent form which is very distinct from other existing methods. Here, all the variables are considered to be triangular fuzzy numbers. Utilizing a parametric form of triangular fuzzy numbers such as left fuzziness index, right fuzziness index, modal value and by using proposed algorithm, we obtained the initial basic fuzzy feasible solution to the problem without changing to its crisp equivalent form. To obtain the solution, we use fuzzy ranking dependent on left and right spreads at some S-level of fuzzy numbers and fuzzy arithmetic dependent on both location index function and fuzziness index function. We further discuss the more for less solution by obtaining shadow prices to the problem. The more for less situation occurs when we raise the supplies, and demand quantities may cause to reduce the optimal transportation cost. The more for less paradox is very useful to a manager in dynamic, for example expanding an item house loading level or plant age breaking point and publicizing attempts to build request at explicit markets. A numerical example is solved to show the adequacy of the proposed algorithm. [ABSTRACT FROM AUTHOR]