Disambiguation Enabled Linear Discriminant Analysis for Partial Label Dimensionality Reduction

As an emerging weakly supervised learning framework, partial label learning considers inaccurate supervision where each training example is associated with multiple candidate labels among which only one is valid. In this article, a first attempt toward employing dimensionality reduction to help improve the generalization performance of partial label learning system is investigated. Specifically, the popular linear discriminant analysis (LDA) techniques are endowed with the ability of dealing with partial label training examples. To tackle the challenge of unknown ground-truth labeling information, a novel learning approach named Delin is proposed which alternates between LDA dimensionality reduction and candidate label disambiguation based on estimated labeling confidences over candidate labels. On one hand, the (kernelized) projection matrix of LDA is optimized by utilizing disambiguation-guided labeling confidences. On the other hand, the labeling confidences are disambiguated by resorting to k NN aggregation in the LDA-induced feature space. Extensive experiments over a broad range of partial label datasets clearly validate the effectiveness of Delin in improving the generalization performance of well-established partial label learning algorithms.

Immunological Classification of Pancreatic Carcinomas to Identify Immune Index and Provide a Strategy for Patient Stratification

BackgroundCancer immunotherapy has produced significant positive clinical effects in a variety of tumor types. However, pancreatic ductal adenocarcinoma (PDAC) is widely considered to be a “cold” cancer with poor immunogenicity. Our aim is to determine the detailed immune features of PDAC to seek new treatment strategies.MethodsThe immune cell abundance of PDAC patients was evaluated with the single-sample gene set enrichment analysis (ssGSEA) using 119 immune gene signatures. Based on these data, patients were classified into different immune subtypes (ISs) according to immune gene signatures. We analyzed their response patterns to immunotherapy in the datasets, then established an immune index to reflect the different degrees of immune infiltration through linear discriminant analysis (LDA). Finally, potential prognostic markers associated with the immune index were identified based on weighted correlation network analysis (WGCNA) that was functionally validated in vitro.ResultsThree ISs were identified in PDAC, of which IS3 had the best prognosis across all three cohorts. The different expressions of immune profiles among the three ISs indicated a distinct responsiveness to immunotherapies in PDAC subtypes. By calculating the immune index, we found that the IS3 represented higher immune infiltration, while IS1 represented lower immune infiltration. Among the investigated signatures, we identified ZNF185, FANCG, and CSTF2 as risk factors associated with immune index that could potentially facilitate diagnosis and could be therapeutic target markers in PDAC patients.ConclusionsOur findings identified immunologic subtypes of PDAC with distinct prognostic implications, which allowed us to establish an immune index to represent the immune infiltration in each subtype. These results show the importance of continuing investigation of immunotherapy and will allow clinical workers to personalized treatment more effectively in PDAC patients.

Electroencephalographic Evidence for Individual Neural Inertia in Mice That Decreases With Time

Previous studies have demonstrated that the brain has an intrinsic resistance to changes in arousal state. This resistance is most easily measured at the population level in the setting of general anesthesia and has been termed neural inertia. To date, no study has attempted to determine neural inertia in individuals. We hypothesize that individuals with markedly increased or decreased neural inertia might be at increased risk for complications related to state transitions, from awareness under anesthesia, to delayed emergence or confusion/impairment after emergence. Hence, an improved theoretical and practical understanding of neural inertia may have the potential to identify individuals at increased risk for these complications. This study was designed to explicitly measure neural inertia in individuals and empirically test the stochastic model of neural inertia using spectral analysis of the murine EEG. EEG was measured after induction of and emergence from isoflurane administered near the EC50 dose for loss of righting in genetically inbred mice on a timescale that minimizes pharmacokinetic confounds. Neural inertia was assessed by employing classifiers constructed using linear discriminant or supervised machine learning methods to determine if features of EEG spectra reliably demonstrate path dependence at steady-state anesthesia. We also report the existence of neural inertia at the individual level, as well as the population level, and that neural inertia decreases over time, providing direct empirical evidence supporting the predictions of the stochastic model of neural inertia.

Viral dysbiosis in children with new-onset celiac disease

Viruses are common components of the intestinal microbiome, modulating host bacterial metabolism and interacting with the immune system, with a possible role in the pathogenesis of immune-mediated diseases such as celiac disease (CeD). The objective of this study was to characterize the virome profile in children with new-onset CeD. We used metagenomic analysis of viral DNA in mucosal and fecal samples from children with CeD and controls and performed sequencing using the Nextera XT library preparation kit. Abundance log2 fold changes were calculated using differential expression and linear discriminant effect size. Shannon alpha and Bray–Curtis beta diversity were determined. A total of 40 children with CeD and 39 controls were included. We found viral dysbiosis in both fecal and mucosal samples. Examples of significantly more abundant species in fecal samples of children with CeD included Human polyomavirus 2, Enterobacteria phage mEpX1, and Enterobacteria phage mEpX2; whereas less abundant species included Lactococcus phages ul36 and Streptococcus phage Abc2. In mucosal samples however, no species were significantly associated with CeD. Shannon alpha diversity was not significantly different between CeD and non-CeD groups and Bray–Curtis beta diversity showed no significant separation between CeD and non-CeD samples in either mucosal or stool samples, whereas separation was clear in all samples. We identified significant viral dysbiosis in children with CeD, suggesting a potential role in the pathogenesis of CeD indicating the need for further studies.

How Molecular Topology Can Help in Amyotrophic Lateral Sclerosis (ALS) Drug Development: A Revolutionary Paradigm for a Merciless Disease

Even if amyotrophic lateral sclerosis is still considered an orphan disease to date, its prevalence among the population is growing fast. Despite the efforts made by researchers and pharmaceutical companies, the cryptic information related to the biological and physiological onset mechanisms, as well as the complexity in identifying specific pharmacological targets, make it almost impossible to find effective treatments. Furthermore, because of complex ethical and economic aspects, it is usually hard to find all the necessary resources when searching for drugs for new orphan diseases. In this context, computational methods, based either on receptors or ligands, share the capability to improve the success rate when searching and selecting potential candidates for further experimentation and, consequently, reduce the number of resources and time taken when delivering a new drug to the market. In the present work, a computational strategy based on Molecular Topology, a mathematical paradigm capable of relating the chemical structure of a molecule to a specific biological or pharmacological property by means of numbers, is presented. The result was the creation of a reliable and accessible tool to help during the early in silico stages in the identification and repositioning of potential hits for ALS treatment, which can also apply to other orphan diseases. Considering that further computational and experimental results will be required for the final identification of viable hits, three linear discriminant equations combined with molecular docking simulations on specific proteins involved in ALS are reported, along with virtual screening of the Drugbank database as a practical example. In this particular case, as reported, a clinical trial has been already started for one of the drugs proposed in the present study.

Sialidase Activity in the Cervicovaginal Fluid Is Associated With Changes in Bacterial Components of Lactobacillus-Deprived Microbiota

IntroductionSialidase activity in the cervicovaginal fluid (CVF) is associated with microscopic findings of bacterial vaginosis (BV). Sequencing of bacterial 16S rRNA gene in vaginal samples has revealed that the majority of microscopic BV cases fit into vaginal community-state type IV (CST IV), which was recently named “molecular-BV.” Bacterial vaginosis-associated bacterial species, such as Gardnerella spp., may act as sources of CVF sialidases. These hydrolases lead to impairment of local immunity and enable bacterial adhesion to epithelial and biofilm formation. However, the impact of CVL sialidase on microbiota components and diversity remains unknown.ObjectiveTo assess if CVF sialidase activity is associated with changes in bacterial components of CST IV.MethodsOne hundred forty women were cross-sectionally enrolled. The presence of molecular-BV (CST IV) was assessed by V3–V4 16S rRNA sequencing (Illumina). Fluorometric assays were performed using 2-(4-methylumbelliferyl)-α-D-N-acetylneuraminic acid (MUAN) for measuring sialidase activity in CVF samples. Linear discriminant analysis effect size (LEfSe) was performed to identify the differently enriched bacterial taxa in molecular-BV according to the status of CVF sialidase activity.ResultsForty-four participants (31.4%) had molecular-BV, of which 30 (68.2%) had sialidase activity at detectable levels. A total of 24 bacterial taxa were enriched in the presence of sialidase activity, while just two taxa were enriched in sialidase-negative samples.ConclusionSialidase activity in molecular-BV is associated with changes in bacterial components of the local microbiome. This association should be further investigated, since it may result in diminished local defenses against pathogens.

Interclutch variability in egg characteristics in two species of rail: Is maternal identity encoded in eggshell patterns?

Maternal signatures are present in the eggs of some birds, but quantifying interclutch variability within populations remains challenging. Maternal assignment of eggs with distinctive appearances could be used to non-invasively identify renesting females, including hens returning among years, as well as to identify cases of conspecific brood parasitism. We explored whether King Rail (Rallus elegans) eggs with shared maternity could be matched based on eggshell pattern. We used NaturePatternMatch (NPM) software to match egg images taken in the field in conjunction with spatial and temporal data on nests. Since we had only a small number of marked breeders, we analyzed similar clutch images from a study of Eurasian Common Moorhens (Gallinula chloropus chloropus) with color-banded breeders for which parentage at many nests had been verified genetically to validate the method. We ran 66 King Rail clutches (n = 338 eggs) and 58 Common Moorhen clutches (n = 364 eggs) through NPM. We performed non-metric multidimensional scaling and permutational analysis of variance using the best egg match output from NPM. We also explored whether eggs could be grouped by clutch using a combination of egg dimensions and pattern data derived from NPM using linear discriminant analyses. We then scrutinized specific matches returned by NPM for King Rail eggs to determine whether multiple matches between the same clutches might reveal maternity among nests and inform our understanding of female laying behavior. To do this, we ran separate NPM analyses for clutches photographed over several years from two spatially distant parts of the site. With these narrower datasets, we were able to identify four instances where hens likely returned to breed among years, four likely cases of conspecific brood parasitism, and a within-season re-nesting attempt. Thus, the matching output was helpful in identifying congruent egg patterns among clutches when used in conjunction with spatial and temporal data, revealing previously unrecognized site fidelity, within-season movements, and reproductive interference by breeding females. Egg pattern data in combination with nest mapping can be used to inform our understanding of female reproductive effort, success, and longevity in King Rails. These methods may also be applied to other secretive birds and species of conservation concern.

A Novel Motion Recognition Method Based on Force Myography of Dynamic Muscle Contractions

Surface electromyogram-based pattern recognition (sEMG-PR) has been considered as the most promising method to control multifunctional prostheses for decades. However, the commercial applications of sEMG-PR in prosthetic control is still limited due to the ambient noise and impedance variation between electrodes and skin surface. In order to reduce these issues, a force-myography-based pattern recognition method was proposed. In this method, a type of polymer-based flexible film sensors, the piezoelectrets, were used to record the rate of stress change (RSC) signals on the muscle surface of eight able-bodied subjects for six hand motions. Thirteen time domain features and four classification algorithms of linear discriminant analysis (LDA), K-nearest neighbor (KNN), artificial neural network (ANN), and support vector machine (SVM) were adopted to decode the RSC signals of different motion classes. In addition, the optimal feature set, classifier, and analysis window length were investigated systematically. Results showed that the average classification accuracy was 95.5 ± 2.2% by using the feature combination of root mean square (RMS) and waveform length (WL) for the classifier of KNN, and the analysis window length of 300 ms was found to obtain the best classification performance. Moreover, the robustness of the proposed method was investigated, and the classification accuracies were observed above 90% even when the white noise ratio increased to 50%. The work of this study demonstrated the effectiveness of RSC-based pattern recognition method for motion classification, and it would provide an alternative approach for the control of multifunctional prostheses.