Benchmarking feature selection and feature extraction methods to improve the performances of machine-learning algorithms for patient classification using metabolomics biomedical data

Objective

Classification tasks are an open challenge in the field of biomedicine. While several machine-learning techniques exist to accomplish this objective, several peculiarities associated with biomedical data, especially when it comes to omics measurements, prevent their use or good performance achievements. Omics approaches aim to understand a complex biological system through systematic analysis of its content at the molecular level. On the other hand, omics data are heterogeneous, sparse and affected by the classical "curse of dimensionality" problem, i.e. having much fewer observation, samples (n) than omics features (p). Furthermore, a major problem with multi-omics data is the imbalance either at the class or feature level. The objective of this work is to study whether feature extraction and/or feature selection techniques can improve the performances of classification machine-learning algorithms on omics measurements.

Methods

Among all omics, metabolomics has emerged as a powerful tool in cancer research, facilitating a deeper understanding of the complex metabolic landscape associated with tumorigenesis and tumor progression. Thus, we selected three publicly available metabolomics datasets, and we applied several feature extraction techniques both linear and non-linear, coupled or not with feature selection methods, and evaluated the performances regarding patient classification in the different configurations for the three datasets.

Results

We provide general workflow and guidelines on when to use those techniques depending on the characteristics of the data available. To further test the extension of our approach to other omics data, we have included a transcriptomics and a proteomics data. Overall, for all datasets, we showed that applying supervised feature selection improves the performances of feature extraction methods for classification purposes. Scripts used to perform all analyses are available at: https://github.com/Plant-Net/Metabolomic_project/.

Multi-objective context-guided consensus of a massive array of techniques for the inference of Gene Regulatory Networks

BACKGROUND AND OBJECTIVE

Gene Regulatory Network (GRN) inference is a fundamental task in biology and medicine, as it enables a deeper understanding of the intricate mechanisms of gene expression present in organisms. This bioinformatics problem has been addressed in the literature through multiple computational approaches. Techniques developed for inferring from expression data have employed Bayesian networks, ordinary differential equations (ODEs), machine learning, information theory measures and neural networks, among others. The diversity of implementations and their respective customization have led to the emergence of many tools and multiple specialized domains derived from them, understood as subsets of networks with specific characteristics that are challenging to detect a priori. This specialization has introduced significant uncertainty when choosing the most appropriate technique for a particular dataset. This proposal, named MO-GENECI, builds upon the basic idea of the previous proposal GENECI and optimizes consensus among different inference techniques, through a carefully refined multi-objective evolutionary algorithm guided by various objective functions, linked to the biological context at hand.

METHODS

MO-GENECI has been tested on an extensive and diverse academic benchmark of 106 gene regulatory networks from multiple sources and sizes. The evaluation of MO-GENECI compared its performance to individual techniques using key metrics (AUROC and AUPR) for gene regulatory network inference. Friedman's statistical ranking provided an ordered classification, followed by non-parametric Holm tests to determine statistical significance.

RESULTS

MO-GENECI's Pareto front approximation facilitates easy selection of an appropriate solution based on generic input data characteristics. The best solution consistently emerged as the winner in all statistical tests, and in many cases, the median precision solution showed no statistically significant difference compared to the winner.

CONCLUSIONS

MO-GENECI has not only demonstrated achieving more accurate results than individual techniques, but has also overcome the uncertainty associated with the initial choice due to its flexibility and adaptability. It is shown intelligently to select the most suitable techniques for each case. The source code is hosted in a public repository at GitHub under MIT license: https://github.com/AdrianSeguraOrtiz/MO-GENECI. Moreover, to facilitate its installation and use, the software associated with this implementation has been encapsulated in a Python package available at PyPI: https://pypi.org/project/geneci/.

Exploring gene regulation and biological processes in insects: Insights from omics data using gene regulatory network models

Gene regulatory network (GRN) comprises complicated yet intertwined gene-regulator relationships. Understanding the GRN dynamics will unravel the complexity behind the observed gene expressions. Insect gene regulation is often complicated due to their complex life cycles and diverse ecological adaptations. The main interest of this review is to have an update on the current mathematical modelling methods of GRNs to explain insect science. Several popular GRN architecture models are discussed, together with examples of applications in insect science. In the last part of this review, each model is compared from different aspects, including network scalability, computation complexity, robustness to noise and biological relevancy.

Three-layer heterogeneous network based on the integration of CircRNA information for MiRNA-disease association prediction

Increasing research has shown that the abnormal expression of microRNA (miRNA) is associated with many complex diseases. However, biological experiments have many limitations in identifying the potential disease-miRNA associations. Therefore, we developed a computational model of Three-Layer Heterogeneous Network based on the Integration of CircRNA information for MiRNA-Disease Association prediction (TLHNICMDA). In the model, a disease-miRNA-circRNA heterogeneous network is built by known disease-miRNA associations, known miRNA-circRNA interactions, disease similarity, miRNA similarity, and circRNA similarity. Then, the potential disease-miRNA associations are identified by an update algorithm based on the global network. Finally, based on global and local leave-one-out cross validation (LOOCV), the values of AUCs in TLHNICMDA are 0.8795 and 0.7774. Moreover, the mean and standard deviation of AUC in 5-fold cross-validations is 0.8777+/-0.0010. Especially, the two types of case studies illustrated the usefulness of TLHNICMDA in predicting disease-miRNA interactions.

Predicting lysine methylation sites using a convolutional neural network

Protein lysine methylation is a particular type of post translational modification that plays an important role in both histone and non-histone function regulation in proteins. Deregulation caused by lysine methyltransferases has been identified as the cause of several diseases including cancer as well as both mental and developmental disorders. Identifying lysine methylation sites is a critical step in both early diagnosis and drug design. This study proposes a new Machine Learning method called CNN-Meth for predicting lysine methylation sites using a convolutional neural network (CNN). Our model is trained using evolutionary, structural, and physicochemical-based presentation along with binary encoding. Unlike previous studies, instead of extracting handcrafted features, we use CNN to automatically extract features from different presentations of amino acids to avoid information loss. Automated feature extraction from these representations of amino acids as well as CNN as a classifier have never been used for this problem. Our results demonstrate that CNN-Meth can significantly outperform previous methods for predicting methylation sites. It achieves 96.0%, 85.1%, 96.4%, and 0.65 in terms of Accuracy, Sensitivity, Specificity, and Matthew's Correlation Coefficient (MCC), respectively. CNN-Meth and its source code are publicly available at https://github.com/MLBC-lab/CNN-Meth.

Prediction of miRNAs and diseases association based on sparse autoencoder and MLP

Introduction: MicroRNAs (miRNAs) are small and non-coding RNA molecules which have multiple important regulatory roles within cells. With the deepening research on miRNAs, more and more researches show that the abnormal expression of miRNAs is closely related to various diseases. The relationship between miRNAs and diseases is crucial for discovering the pathogenesis of diseases and exploring new treatment methods. Methods: Therefore, we propose a new sparse autoencoder and MLP method (SPALP) to predict the association between miRNAs and diseases. In this study, we adopt advanced deep learning technologies, including sparse autoencoder and multi-layer perceptron (MLP), to improve the accuracy of predicting miRNA-disease associations. Firstly, the SPALP model uses a sparse autoencoder to perform feature learning and extract the initial features of miRNAs and diseases separately, obtaining the latent features of miRNAs and diseases. Then, the latent features combine miRNAs functional similarity data with diseases semantic similarity data to construct comprehensive miRNAs-diseases datasets. Subsequently, the MLP model can predict the unknown association among miRNAs and diseases. Result: To verify the performance of our model, we set up several comparative experiments. The experimental results show that, compared with traditional methods and other deep learning prediction methods, our method has significantly improved the accuracy of predicting miRNAs-disease associations, with 94.61% accuracy and 0.9859 AUC value. Finally, we conducted case study of SPALP model. We predicted the top 30 miRNAs that might be related to Lupus Erythematosus, Ecute Myeloid Leukemia, Cardiovascular, Stroke, Diabetes Mellitus five elderly diseases and validated that 27, 29, 29, 30, and 30 of the top 30 are indeed associated. Discussion: The SPALP approach introduced in this study is adept at forecasting the links between miRNAs and diseases, addressing the complexities of analyzing extensive bioinformatics datasets and enriching the comprehension contribution to disease progression of miRNAs.

Prediction of protein N-terminal acetylation modification sites based on CNN-BiLSTM-attention model

N-terminal acetylation is one of the most common and important post-translational modifications (PTM) of eukaryotic proteins. PTM plays a crucial role in various cellular processes and disease pathogenesis. Thus, the accurate identification of N-terminal acetylation modifications is important to gain insight into cellular processes and other possible functional mechanisms. Although some algorithmic models have been proposed, most have been developed based on traditional machine learning algorithms and small training datasets. Their practical applications are limited. Nevertheless, deep learning algorithmic models are better at handling high-throughput and complex data. In this study, DeepCBA, a model based on the hybrid framework of convolutional neural network (CNN), bidirectional long short-term memory network (BiLSTM), and attention mechanism deep learning, was constructed to detect the N-terminal acetylation sites. The DeepCBA was built as follows: First, a benchmark dataset was generated by selecting low-redundant protein sequences from the Uniport database and further reducing the redundancy of the protein sequences using the CD-HIT tool. Subsequently, based on the skip-gram model in the word2vec algorithm, tripeptide word vector features were generated on the benchmark dataset. Finally, the CNN, BiLSTM, and attention mechanism were combined, and the tripeptide word vector features were fed into the stacked model for multiple rounds of training. The model performed excellently on independent dataset test, with accuracy and area under the curve of 80.51% and 87.36%, respectively. Altogether, DeepCBA achieved superior performance compared with the baseline model, and significantly outperformed most existing predictors. Additionally, our model can be used to identify disease loci and drug targets.

A Survey of Deep Learning for Detecting miRNA- Disease Associations: Databases, Computational Methods, Challenges, and Future Directions

MicroRNAs (miRNAs) are an important class of non-coding RNAs that play an essential role in the occurrence and development of various diseases. Identifying the potential miRNA-disease associations (MDAs) can be beneficial in understanding disease pathogenesis. Traditional laboratory experiments are expensive and time-consuming. Computational models have enabled systematic large-scale prediction of potential MDAs, greatly improving the research efficiency. With recent advances in deep learning, it has become an attractive and powerful technique for uncovering novel MDAs. Consequently, numerous MDA prediction methods based on deep learning have emerged. In this review, we first summarize publicly available databases related to miRNAs and diseases for MDA prediction. Next, we outline commonly used miRNA and disease similarity calculation and integration methods. Then, we comprehensively review the 48 existing deep learning-based MDA computation methods, categorizing them into classical deep learning and graph neural network-based techniques. Subsequently, we investigate the evaluation methods and metrics that are frequently used to assess MDA prediction performance. Finally, we discuss the performance trends of different computational methods, point out some problems in current research, and propose 9 potential future research directions. Data resources and recent advances in MDA prediction methods are summarized in the GitHub repository https://github.com/sheng-n/DL-miRNA-disease-association-methods.

Machine Learning Strategies in MicroRNA Research: Bridging Genome to Phenome

MicroRNAs (miRNAs) have emerged as a prominent layer of regulation of gene expression. This article offers the salient and current aspects of machine learning (ML) tools and approaches from genome to phenome in miRNA research. First, we underline that the complexity in the analysis of miRNA function ranges from their modes of biogenesis to the target diversity in diverse biological conditions. Therefore, it is imperative to first ascertain the miRNA coding potential of genomes and understand the regulatory mechanisms of their expression. This knowledge enables the efficient classification of miRNA precursors and the identification of their mature forms and respective target genes. Second, and because one miRNA can target multiple mRNAs and vice versa, another challenge is the assessment of the miRNA-mRNA target interaction network. Furthermore, long-noncoding RNA (lncRNA)and circular RNAs (circRNAs) also contribute to this complexity. ML has been used to tackle these challenges at the high-dimensional data level. The present expert review covers more than 100 tools adopting various ML approaches pertaining to, for example, (1) miRNA promoter prediction, (2) precursor classification, (3) mature miRNA prediction, (4) miRNA target prediction, (5) miRNA- lncRNA and miRNA-circRNA interactions, (6) miRNA-mRNA expression profiling, (7) miRNA regulatory module detection, (8) miRNA-disease association, and (9) miRNA essentiality prediction. Taken together, we unpack, critically examine, and highlight the cutting-edge synergy of ML approaches and miRNA research so as to develop a dynamic and microlevel understanding of human health and diseases.

The impact of unmet treatment need on oral health related quality of life: a questionnaire survey

BACKGROUND

Based on the present global burden of oral diseases, unmet dental needs affect a more significant population worldwide. It is characterised by the need for dental care but receiving delayed or no care. The contributing factors include lack of knowledge about oral health, its consequences, and the availability of dental services. We need to find out the scale of the problem of unmet dental needs for the south Indian population. Therefore, the objective was to determine the relationship between the presence of oral disease and the quality of life-related to oral health using the OHIP-14 tool.

METHODS

The unmet dental requirements of the south Indian population were determined using a cross-sectional questionnaire survey. Close-ended questions were used to obtain data from two investigators trained to record the answers from the patients. The data was collected using the OHIP-14 questionnaire, which consists of 14 items divided into seven domains with two questions each. Physical pain, psychological impairment, physical disability, psychological disability, social disability, and disability were all considered. An additional analysis of artificial neural network (ANN) was done.

RESULTS

The response rate was 100 per cent. N = 1029 people replied to the questionnaire about their unmet dental needs. N = 497 (48.3%) were men, whereas N = 532 (51.7%) were women. The average age was 31.7811.72. As their current occupation, most of the included subjects (60.1%) were students. The respondents had no known personal habits and a mixed diet (94.93%). The average BMI was 24.022.59 (14-30.9). OHIP was present in 62.3% of the population. The average OHIP-14 severity score was 10.97. (8.54). The severity and degree of unmet dental need were substantial (p0.01) due to pain in the mouth/teeth/gums, malocclusion, and gum bleeding. The most common OHIP-14 domains affected by unmet oral needs were psychological discomfort, psychological limitation, social limitation, and feeling handicapped. The analysis of ANN revealed that high OHIP scores were primarily attributed to dental caries, poor oral health, and dental aesthetics.

CONCLUSION

The severity and degree of unmet dental needs were significant among the south Indian population. The most common oral health status that impacted OHIP-14 domains were pain, malocclusion, and bleeding gums. These patients were significantly impacted by psychological discomfort and social limitations and felt handicapped.

Understanding wound healing in obesity

Obesity has become more prevalent in the global population. It is associated with the development of several diseases including diabetes mellitus, coronary heart disease, and metabolic syndrome. There are a multitude of factors impacted by obesity that may contribute to poor wound healing outcomes. With millions worldwide classified as obese, it is imperative to understand wound healing in these patients. Despite advances in the understanding of wound healing in both healthy and diabetic populations, much is unknown about wound healing in obese patients. This review examines the impact of obesity on wound healing and several animal models that may be used to broaden our understanding in this area. As a growing portion of the population identifies as obese, understanding the underlying mechanisms and how to overcome poor wound healing is of the utmost importance.

Maternal Acylcarnitine Disruption as a Potential Predictor of Preterm Birth in Primigravida: A Preliminary Investigation

Preterm birth, defined as any birth before 37 weeks of completed gestation, poses adverse health risks to both mothers and infants. Despite preterm birth being associated with several risk factors, its relationship to maternal metabolism remains unclear, especially in first-time mothers. Aims of the present study were to identify maternal metabolic disruptions associated with preterm birth and to evaluate their predictive potentials. Blood was collected, and the serum harvested from the mothers of 24 preterm and 42 term births at 28-32 weeks gestation (onset of the 3rd trimester). Serum samples were assayed by untargeted metabolomic analyses via liquid chromatography/mass spectrometry (QTOF-LC/MS). Metabolites were annotated by inputting the observed mass-to-charge ratio into the Human Metabolome Database (HMDB). Analysis of 181 identified metabolites by PLS-DA modeling using SIMCA (v17) showed reasonable separation between the two groups (CV-ANOVA, p = 0.02). Further statistical analysis revealed lower serum levels of various acyl carnitines and amino acid metabolites in preterm mothers. Butenylcarnitine (C4:1), a short-chain acylcarnitine, was found to be the most predictive of preterm birth (AUROC = 0.73, [CI] 0.60-0.86). These observations, in conjuncture with past literature, reveal disruptions in fatty acid oxidation and energy metabolism in preterm primigravida. While these findings require validation, they reflect altered metabolic pathways that may be predictive of preterm delivery in primigravida.

Deep phenotyping of post-infectious myalgic encephalomyelitis/chronic fatigue syndrome

Post-infectious myalgic encephalomyelitis/chronic fatigue syndrome (PI-ME/CFS) is a disabling disorder, yet the clinical phenotype is poorly defined, the pathophysiology is unknown, and no disease-modifying treatments are available. We used rigorous criteria to recruit PI-ME/CFS participants with matched controls to conduct deep phenotyping. Among the many physical and cognitive complaints, one defining feature of PI-ME/CFS was an alteration of effort preference, rather than physical or central fatigue, due to dysfunction of integrative brain regions potentially associated with central catechol pathway dysregulation, with consequences on autonomic functioning and physical conditioning. Immune profiling suggested chronic antigenic stimulation with increase in naïve and decrease in switched memory B-cells. Alterations in gene expression profiles of peripheral blood mononuclear cells and metabolic pathways were consistent with cellular phenotypic studies and demonstrated differences according to sex. Together these clinical abnormalities and biomarker differences provide unique insight into the underlying pathophysiology of PI-ME/CFS, which may guide future intervention.

Intra-subject enveloped multilayer fuzzy sample compression for speech diagnosis of Parkinson's disease

Machine learning-based Parkinson's disease (PD) speech diagnosis is a current research hotspot. However, existing methods use each corpus sample as the base unit for modeling. Since different corpus samples within the same subject have different sensitive speech features, it is difficult to obtain unified and stable sensitive speech features (diagnostic markers) that reflect the pathology of the whole subject. Therefore, this study aims at compressing the corpus samples within the subject to facilitate the search for diagnostic markers with high diagnostic accuracy. A two-step sample compression module (TSCM) can solve the problem above. It includes two major parts: sample pruning module (SPM) and sample fuzzy clustering mechanism (SFCMD). Based on stacking multiple TSCMs, a multilayer sample compression module (MSCM) is formed to obtain multilayer compression samples. After that, simultaneous sample/feature selection mechanism (SS/FSM) is designed for feature selection. Based on the multilayer compression samples processed by MSCM and SS/FSM, a novel ensemble learning algorithm (EMSFE) is designed with sparse fusion ensemble learning mechanism (SFELM). The proposed EMSFE is validated by visualization of extracted features and performance comparison with related algorithms. The experimental results show that the proposed algorithm can effectively extract the stable diagnostic markers by compressing the corpus samples within the subject. Furthermore, based on LOSO cross validation, the proposed algorithm with extreme learning machine (ELM) classifier can achieve the accuracy of 92.5%, 93.75% and 91.67% on three datasets, respectively. The proposed EMSFE can extract unified and stable sensitive features that accurately reflect the overall pathology of the subject, which can better meet the requirements of clinical applications.The code and datasets can be found in: https://github.com/wywwwww/EMSFE-supplementary-material.git.

FastKnock: an efficient next-generation approach to identify all knockout strategies for strain optimization

Overproduction of desired native or nonnative biochemical(s) in (micro)organisms can be achieved through metabolic engineering. Appropriate rewiring of cell metabolism is performed by making rational changes such as insertion, up-/down-regulation and knockout of genes and consequently metabolic reactions. Finding appropriate targets (including proper sets of reactions to be knocked out) for metabolic engineering to design optimal production strains has been the goal of a number of computational algorithms. We developed FastKnock, an efficient next-generation algorithm for identifying all possible knockout strategies (with a predefined maximum number of reaction deletions) for the growth-coupled overproduction of biochemical(s) of interest. We achieve this by developing a special depth-first traversal algorithm that allows us to prune the search space significantly. This leads to a drastic reduction in execution time. We evaluate the performance of the FastKnock algorithm using various Escherichia coli genome-scale metabolic models in different conditions (minimal and rich mediums) for the overproduction of a number of desired metabolites. FastKnock efficiently prunes the search space to less than 0.2% for quadruple- and 0.02% for quintuple-reaction knockouts. Compared to the classic approaches such as OptKnock and the state-of-the-art techniques such as MCSEnumerator methods, FastKnock found many more beneficial and important practical solutions. The availability of all the solutions provides the opportunity to further characterize, rank and select the most appropriate intervention strategy based on any desired evaluation index. Our implementation of the FastKnock method in Python is publicly available at https://github.com/leilahsn/FastKnock .

Distributed multi-objective optimization for SNP-SNP interaction detection

The detection of complex interactions between single nucleotide polymorphisms (SNPs) plays a vital role in genome-wide association analysis (GWAS). The multi-objective evolutionary algorithm is a promising technique for SNP-SNP interaction detection. However, as the scale of SNP data further increases, the exponentially growing search space gradually becomes the dominant factor, causing evolutionary algorithm (EA)-based approaches to fall into local optima. In addition, multi-objective genetic operations consume significant amounts of time and computational resources. To this end, this study proposes a distributed multi-objective evolutionary framework (DM-EF) to identify SNP-SNP interactions on large-scale datasets. DM-EF first partitions the entire search space into several subspaces based on a space-partitioning strategy, which is nondestructive because it guarantees that each feasible solution is assigned to a specific subspace. Thereafter, each subspace is optimized using a multi-objective EA optimizer, and all subspaces are optimized in parallel. A decomposition-based multi-objective firework optimizer (DCFWA) with several problem-guided operators was designed. Finally, the final output is selected from the Pareto-optimal solutions in the historical search of each subspace. DM-EF avoids the preference for a single objective function, handles the heavy computational burden, and enhances the diversity of the population to avoid local optima. Notably, DM-EF is load-balanced and scalable because it can flexibly partition the space according to the number of available computational nodes and problem size. Experiments on both artificial and real-world datasets demonstrate that the proposed method significantly improves the search speed and accuracy.

Global protein-protein interaction networks in yeast saccharomyces cerevisiae and helicobacter pylori

Understanding many biological processes relies heavily on accurately predicting protein-protein interactions (PPIs). In this study, we propose a novel method for predicting PPIs that is based on LogitBoost with a binary bat feature selection algorithm. Our approach involves the extraction of an initial feature vector by combining pseudo amino acid composition (PseAAC), pseudo-position-specific scoring matrix (PsePSSM), reduced sequence and index-vectors (RSIV), and autocorrelation descriptor (AD). Subsequently, a binary bat algorithm is applied to eliminate redundant features, and the resulting optimal features are fed into the LogitBoost classifier for the identification of PPIs. To evaluate the proposed method, we test it on two databases, Saccharomyces cerevisiae and Helicobacter pylori, using 10-fold cross-validation, and achieve accuracies of 94.39% and 97.89%, respectively. Our results showcase the significant potential of our pipeline in accurately predicting protein-protein interactions (PPIs), thereby offering a valuable resource to the scientific research community.

Poor Accuracy of Manually Derived Head Computed Tomography Parameters in Predicting Intracranial Hypertension After Nontraumatic Intracranial Hemorrhage

BACKGROUND

The utility of head computed tomography (CT) in predicting elevated intracranial pressure (ICP) is known to be limited in traumatic brain injury; however, few data exist in patients with spontaneous intracranial hemorrhage.

METHODS

We conducted a retrospective review of prospectively collected data in patients with nontraumatic intracranial hemorrhage (subarachnoid hemorrhage [SAH] or intraparenchymal hemorrhage [IPH]) who underwent external ventricular drain (EVD) placement. Head CT scans performed immediately prior to EVD placement were quantitatively reviewed for features suggestive of elevated ICP, including temporal horn diameter, bicaudate index, basal cistern effacement, midline shift, and global cerebral edema. The modified Fisher score (mFS), intraventricular hemorrhage score, and IPH volume were also measured, as applicable. We calculated the accuracy, positive predictive value (PPV), and negative predictive value (NPV) of these radiographic features for the coprimary outcomes of elevated ICP (> 20 mm Hg) at the time of EVD placement and at any time during the hospital stay. Multivariable backward stepwise logistic regression analysis was performed to identify significant radiographic factors associated with elevated ICP.

RESULTS

Of 608 patients with intracranial hemorrhages enrolled during the study time frame, 243 (40%) received an EVD and 165 (n = 107 SAH, n = 58 IPH) had a preplacement head CT scan available for rating. Elevated opening pressure and elevated ICP during hospitalization were recorded in 48 of 152 (29%) and 103 of 165 (62%), respectively. The presence of ≥ 1 radiographic feature had only 32% accuracy for identifying elevated opening pressure (PPV 30%, NPV 58%, area under the curve [AUC] 0.537, 95% asymptotic confidence interval [CI] 0.436-0.637, P = 0.466) and 59% accuracy for predicting elevated ICP during hospitalization (PPV 63%, NPV 40%, AUC 0.514, 95% asymptotic CI 0.391-0.638, P = 0.820). There was no significant association between the number of radiographic features and ICP elevation. Head CT scans without any features suggestive of elevated ICP occurred in 25 of 165 (15%) patients. However, 10 of 25 (40%) of these patients had elevated opening pressure, and 15 of 25 (60%) had elevated ICP during their hospital stay. In multivariable models, mFS (adjusted odds ratio [aOR] 1.36, 95% CI 1.10-1.68) and global cerebral edema (aOR 2.93, 95% CI 1.27-6.75) were significantly associated with elevated ICP; however, their accuracies were only 69% and 60%, respectively. All other individual radiographic features had accuracies between 38 and 58% for identifying intracranial hypertension.

CONCLUSIONS

More than 50% of patients with spontaneous intracranial hemorrhage without radiographic features suggestive of elevated ICP actually had ICP > 20 mm Hg during EVD placement or their hospital stay. Morphological head CT findings were only 32% and 59% accurate in identifying elevated opening pressure and ICP elevation during hospitalization, respectively.

Human Serum and Salivary Metabolomes: Diversity and Closeness

Saliva, which contains molecular information that may reflect an individual's health status, has become a valuable tool for discovering biomarkers of oral and general diseases. Due to the high vascularization of the salivary glands, there is a molecular exchange between blood and saliva. However, the composition of saliva is complex and influenced by multiple factors. This study aimed to investigate the possible relationships between the salivary and serum metabolomes to gain a comprehensive view of the metabolic phenotype under physiological conditions. Using 1H-NMR spectroscopy, we obtained the serum metabolite profiles of 20 healthy young individuals and compared them with the metabolomes of parotid, submandibular/sublingual, and whole-saliva samples collected concurrently from the same individuals using multivariate and univariate statistical analysis. Our results show that serum is more concentrated and less variable for most of the shared metabolites than the three saliva types. While we found moderate to strong correlations between serum and saliva concentrations of specific metabolites, saliva is not simply an ultrafiltrate of blood. The intense oral metabolism prevents very strong correlations between serum and salivary concentrations. This study contributes to a better understanding of salivary metabolic composition, which is crucial for utilizing saliva in laboratory diagnostics.

Soft sensor based on Raman spectroscopy for the in-line monitoring of metabolites and polymer quality in the biomanufacturing of polyhydroxyalkanoates

Polyhydroxyalkanoates (PHA) are among the most promising bio-based alternatives to conventional petroleum-based plastics. These biodegradable polyesters can in fact be produced by fermentation from bacteria like Cupriavidus necator, thus reducing the environmental footprint of the manufacturing process. However, ensuring consistent product quality attributes is a major challenge of biomanufacturing. To address this issue, the implementation of real-time monitoring tools is essential to increase process understanding, enable a prompt response to possible process deviations and realize on-line process optimization. In this work, a soft sensor based on in situ Raman spectroscopy was developed and applied to the in-line monitoring of PHA biomanufacturing. This strategy allows the collection of quantitative information directly from the culture broth, without the need for sampling, and at high frequency. In fact, through an optimized multivariate data analysis pipeline, this soft sensor allows monitoring cell dry weight, as well as carbon and nitrogen source concentrations with root mean squared errors (RMSE) equal to 3.71, 7 and 0.03 g/L, respectively. In addition, this tool allows the in-line monitoring of intracellular PHA accumulation, with an RMSE of 14 gPHA/gCells. For the first time, also the number and weight average molecular weights of the polymer produced could be monitored, with RMSE of 8.7E4 and 11.6E4 g/mol, respectively. Overall, this work demonstrates the potential of Raman spectroscopy in the in-line monitoring of biotechnology processes, leading to the simultaneous measurement of several process variables in real time without the need of sampling and labor-intensive sample preparations.

Breast cancer prediction model based on clinical and biochemical characteristics: clinical data from patients with benign and malignant breast tumors from a single center in South China

OBJECTIVE

Breast cancer is the most prevalent cancer and is second leading cause of death from malignancy among women worldwide. In addition to tumor factors, the host characteristics of tumors have been paid more and more attention by the medical community. This study aimed to develop a breast cancer prediction model for the Chinese population using clinical and biochemical characteristics.

METHODS

This is a retrospective study. From 2012 to 2021, we selected 19,751 patients with breast diseases from the Guangdong Hospital of Traditional Chinese Medicine, which included 5660 patients with breast cancer and 14,091 patients with benign breast diseases-75% of patients were randomly assigned to the training group and 25% to the test group using a total of 34 clinical and biochemical characteristics. Significant clinical signs were investigated, and logistic regression with recursive feature elimination (RFE) model was used to develop a prediction model for distinguishing benign from malignant breast diseases. The prediction model's accuracy, precision, sensitivity, specificity, and area under the ROC curve (AUC) were calculated.

RESULTS

Clinical statistics demonstrated that the prediction model comprised 19 clinical characteristics had statistical separability in both the training group and the test group, as well as good sensitivity and prediction.

CONCLUSIONS

This model based on biochemical parameters demonstrates a significant predictive effect for breast cancer and may be useful as a reference for invasive tissue biopsy in patients undergoing BI-RADS 3 and 4A breast imaging.

Current trends of host-pathogen relationship in shrimp infectious disease via computational protein-protein interaction: A bibliometric analysis

Protein-protein interactions (PPIs) are essential for understanding cell physiology in normal and pathological conditions, as they might involve in all cellular processes. PPIs have been widely used to elucidate the pathobiology of human and plant diseases. Therefore, they can also be used to unveil the pathobiology of infectious diseases in shrimp, which is one of the high-risk factors influencing the success or failure of shrimp production. PPI network analysis, specifically host-pathogen PPI (HP-PPI), provides insights into the molecular interactions between the shrimp and pathogens. This review quantitatively analyzed the research trends within this field through bibliometric analysis using specific keywords, countries, authors, organizations, journals, and documents. This analysis has screened 206 records from the Scopus database for determining eligibility, resulting in 179 papers that were retrieved for bibliometric analysis. The analysis revealed that China and Thailand were the driving forces behind this specific field of research and frequently collaborated with the United States. Aquaculture and Diseases of Aquatic Organisms were the prominent sources for publications in this field. The main keywords identified included "white spot syndrome virus," "WSSV," and "shrimp." We discovered that studies on HP-PPI are currently quite scarce. As a result, we further discussed the significance of HP-PPI by highlighting various approaches that have been previously adopted. These findings not only emphasize the importance of HP-PPI but also pave the way for future researchers to explore the pathogenesis of infectious diseases in shrimp. By doing so, preventative measures and enhanced treatment strategies can be identified.

Estimation and update of betweenness centrality with progressive algorithm and shortest paths approximation

Betweenness centrality is one of the key measures of the node importance in a network. However, it is computationally intractable to calculate the exact betweenness centrality of nodes in large-scale networks. To solve this problem, we present an efficient CBCA (Centroids based Betweenness Centrality Approximation) algorithm based on progressive sampling and shortest paths approximation. Our algorithm firstly approximates the shortest paths by generating the network centroids according to the adjacency information entropy of the nodes; then constructs an efficient error estimator using the Monte Carlo Empirical Rademacher averages to determine the sample size which can achieve a balance with accuracy; finally, we present a novel centroid updating strategy based on network density and clustering coefficient, which can effectively reduce the computation burden of updating shortest paths in dynamic networks. The experimental results show that our CBCA algorithm can efficiently output high-quality approximations of the betweenness centrality of a node in large-scale complex networks.

Who is calling? Optimizing source identification from marmoset vocalizations with hierarchical machine learning classifiers

With their highly social nature and complex vocal communication system, marmosets are important models for comparative studies of vocal communication and, eventually, language evolution. However, our knowledge about marmoset vocalizations predominantly originates from playback studies or vocal interactions between dyads, and there is a need to move towards studying group-level communication dynamics. Efficient source identification from marmoset vocalizations is essential for this challenge, and machine learning algorithms (MLAs) can aid it. Here we built a pipeline capable of plentiful feature extraction, meaningful feature selection, and supervised classification of vocalizations of up to 18 marmosets. We optimized the classifier by building a hierarchical MLA that first learned to determine the sex of the source, narrowed down the possible source individuals based on their sex and then determined the source identity. We were able to correctly identify the source individual with high precisions (87.21%-94.42%, depending on call type, and up to 97.79% after the removal of twins from the dataset). We also examine the robustness of identification across varying sample sizes. Our pipeline is a promising tool not only for source identification from marmoset vocalizations but also for analysing vocalizations of other species.

Inferring circadian gene regulatory relationships from gene expression data with a hybrid framework

BACKGROUND

The central biological clock governs numerous facets of mammalian physiology, including sleep, metabolism, and immune system regulation. Understanding gene regulatory relationships is crucial for unravelling the mechanisms that underlie various cellular biological processes. While it is possible to infer circadian gene regulatory relationships from time-series gene expression data, relying solely on correlation-based inference may not provide sufficient information about causation. Moreover, gene expression data often have high dimensions but a limited number of observations, posing challenges in their analysis.

METHODS

In this paper, we introduce a new hybrid framework, referred to as Circadian Gene Regulatory Framework (CGRF), to infer circadian gene regulatory relationships from gene expression data of rats. The framework addresses the challenges of high-dimensional data by combining the fuzzy C-means clustering algorithm with dynamic time warping distance. Through this approach, we efficiently identify the clusters of genes related to the target gene. To determine the significance of genes within a specific cluster, we employ the Wilcoxon signed-rank test. Subsequently, we use a dynamic vector autoregressive method to analyze the selected significant gene expression profiles and reveal directed causal regulatory relationships based on partial correlation.

CONCLUSION

The proposed CGRF framework offers a comprehensive and efficient solution for understanding circadian gene regulation. Circadian gene regulatory relationships are inferred from the gene expression data of rats based on the Aanat target gene. The results show that genes Pde10a, Atp7b, Prok2, Per1, Rhobtb3 and Dclk1 stand out, which have been known to be essential for the regulation of circadian activity. The potential relationships between genes Tspan15, Eprs, Eml5 and Fsbp with a circadian rhythm need further experimental research.

A New Binary Biclustering Algorithm Based on Weight Adjacency Difference Matrix for Analyzing Gene Expression Data

Biclustering algorithms are essential for processing gene expression data. However, to process the dataset, most biclustering algorithms require preprocessing the data matrix into a binary matrix. Regrettably, this type of preprocessing may introduce noise or cause information loss in the binary matrix, which would reduce the biclustering algorithm's ability to effectively obtain the optimal biclusters. In this paper, we propose a new preprocessing method named Mean-Standard Deviation (MSD) to resolve the problem. Additionally, we introduce a new biclustering algorithm called Weight Adjacency Difference Matrix Binary Biclustering (W-AMBB) to effectively process datasets containing overlapping biclusters. The basic idea is to create a weighted adjacency difference matrix by applying weights to a binary matrix that is derived from the data matrix. This allows us to identify genes with significant associations in sample data by efficiently identifying similar genes that respond to specific conditions. Furthermore, the performance of the W-AMBB algorithm was tested on both synthetic and real datasets and compared with other classical biclustering methods. The experiment results demonstrate that the W-AMBB algorithm is significantly more robust than the compared biclustering methods on the synthetic dataset. Additionally, the results of the GO enrichment analysis show that the W-AMBB method possesses biological significance on real datasets.

CFNCM: Collaborative filtering neighborhood-based model for predicting miRNA-disease associations

MicroRNAs have a significant role in the emergence of various human disorders. Consequently, it is essential to understand the existing interactions between miRNAs and diseases, as this will help scientists better study and comprehend the diseases' biological mechanisms. Findings can be employed as biomarkers or drug targets to advance the detection, diagnosis, and treatment of complex human disorders by foretelling possible disease-related miRNAs. This study proposed a computational model for predicting potential miRNA-disease associations called the Collaborative Filtering Neighborhood-based Classification Model (CFNCM), in light of the shortcomings of conventional and biological experiments, which are expensive and time-consuming. The model generated integrated miRNA and disease similarity matrices using the validated associations and miRNA and disease similarity information and used them as the input features for CFNCM. To produce class labels, we first determined the association scores for brand-new pairs using user-based collaborative filtering. With zero as the threshold, the associations with scores >0 were labelled 1, indicating a potential positive association, otherwise, it is marked as 0. Then, we developed classification models using various machine-learning algorithms. By comparison, we discovered that the support vector machine (SVM) produced the best AUC of 0.96 with 10-fold cross-validation through the GridSearchCV technique for identifying optimal parameter values. In addition, the models were evaluated and verified by analyzing the top 50 breast and lung neoplasms-related miRNAs, of which 46 and 47 associations were verified in two authoritative databases, dbDEMC and miR2Disease.

Predicting miRNA-Disease Associations Through Deep Autoencoder With Multiple Kernel Learning

Determining microRNA (miRNA)-disease associations (MDAs) is an integral part in the prevention, diagnosis, and treatment of complex diseases. However, wet experiments to discern MDAs are inefficient and expensive. Hence, the development of reliable and efficient data integrative models for predicting MDAs is of significant meaning. In the present work, a novel deep learning method for predicting MDAs through deep autoencoder with multiple kernel learning (DAEMKL) is presented. Above all, DAEMKL applies multiple kernel learning (MKL) in miRNA space and disease space to construct miRNA similarity network and disease similarity network, respectively. Then, for each disease or miRNA, its feature representation is learned from the miRNA similarity network and disease similarity network via the regression model. After that, the integrated miRNA feature representation and disease feature representation are input into deep autoencoder (DAE). Furthermore, the novel MDAs are predicted through reconstruction error. Ultimately, the AUC results show that DAEMKL achieves outstanding performance. In addition, case studies of three complex diseases further prove that DAEMKL has excellent predictive performance and can discover a large number of underlying MDAs. On the whole, our method DAEMKL is an effective method to identify MDAs.

Genetic interaction in the association between oxidative stress and diabetes in the Spanish population

Oxidative stress (OS) is a relevant intermediate mechanism involved in Type 2 Diabetes Mellitus (T2D) development. To date, the interaction between OS parameters and variations in genes related to T2D has not been analyzed.

AIMS

To study the genetic interaction of genes potentially related to OS levels (redox homeostasis, renin-angiotensin-aldosterone system, endoplasmic stress response, dyslipidemia, obesity and metal transport) and OS and T2D risk in a general population from Spain (the Hortega Study) in relation to the risk of suffering from T2D.

MATERIALS AND METHODS

One thousand five hundred and two adults from the University Hospital Rio Hortega area were studied and 900 single nucleotide polymorphisms (SNPs) from 272 candidate genes were analyzed.

RESULTS

There were no differences in OS levels between cases and controls. Some polymorphisms were associated with T2D and with OS levels. Significant interactions were observed between OS levels and two polymorphisms in relation to T2D presence: rs196904 (ERN1 gene) and rs2410718 (COX7C gene); and between OS levels and haplotypes of the genes: SP2, HFF1A, ILI8R1, EIF2AK2, TXNRD2, PPARA, NDUFS2 and ERN1.

CONCLUSIONS

Our results indicate that genetic variations of the studied genes are associated with OS levels and that their interaction with OS parameters may contribute to the risk of developing T2D in the Spanish general population. These data support the importance of analyzing the influence of OS levels and their interaction with genetic variations in order to establish their real impact in T2D risk. Further studies are required to identify the real relevance of interactions between genetic variations and OS levels and the mechanisms involved in them.

Uncovering the phenotypic consequences of multi-locus imprinting disturbances using genome-wide methylation analysis in genomic imprinting disorders

Imprinted genes are regulated by DNA methylation of imprinted differentially methylated regions (iDMRs). An increasing number of patients with congenital imprinting disorders (IDs) exhibit aberrant methylation at multiple imprinted loci, multi-locus imprinting disturbance (MLID). We examined MLID and its possible impact on clinical features in patients with IDs. Genome-wide DNA methylation analysis (GWMA) using blood leukocyte DNA was performed on 13 patients with Beckwith-Wiedemann syndrome (BWS), two patients with Silver-Russell syndrome (SRS), and four controls. HumanMethylation850 BeadChip analysis for 77 iDMRs (809 CpG sites) identified three patients with BWS and one patient with SRS showing additional hypomethylation, other than the disease-related iDMRs, suggestive of MLID. Two regions were aberrantly methylated in at least two patients with BWS showing MLID: PPIEL locus (chromosome 1: 39559298 to 39559744), and FAM50B locus (chromosome 6: 3849096 to 3849469). All patients with BWS- and SRS-MLID did not show any other clinical characteristics associated with additional involved iDMRs. Exome analysis in three patients with BWS who exhibited multiple hypomethylation did not identify any causative variant related to MLID. This study indicates that a genome-wide approach can unravel MLID in patients with an apparently isolated ID. Patients with MLID showed only clinical features related to the original IDs. Long-term follow-up studies in larger cohorts are warranted to evaluate any possible phenotypic consequences of other disturbed imprinted loci.

A therapeutic epitopes-based vaccine engineering against Salmonella enterica XDR strains for typhoid fever: a Pan-vaccinomics approach

A prevalent food-borne pathogen, Salmonella enterica serotypes Typhi, is responsible for gastrointestinal and systemic infections globally. Salmonella vaccines are the most effective, however, producing a broad-spectrum vaccine remains challenging due to Salmonella's many serotypes. Efforts are urgently required to develop a novel vaccine candidate that can tackle all S. Typhi strains because of their high resistance to multiple kinds of antibiotics (particularly the XDR H58 strain). In this work, we used a computational pangenome-based vaccine design technique on all available (n = 119) S. Typhi reference genomes and identified one TonB-dependent siderophore receptor (WP_001034967.1) as highly conserved and prospective vaccine candidates from the predicted core genome (n = 3,351). The applied pan-proteomics and Immunoinformatic approaches help in the identification of four epitopes that may trigger adequate host body immune responses. Furthermore, the proposed vaccine ensemble demonstrates a stable binding conformation with the examined immunological receptor (HLAs and TRL2/4) and has large interaction energy determined via molecular docking and molecular dynamics simulation techniques. Eventually, an expression vector for the Escherichia. coli K12 strain was constructed from the vaccine sequence. Additional analysis revealed that the vaccine may help to elicit strong immune responses for typhoid infections, however, experimental analysis is required to verify the vaccine's effectiveness based on these results. Moreover, the applied computer-assisted vaccine design may considerably decrease vaccine development costs and speed up the process. The study's findings are intriguing, but they must be evaluated in the experimental labs to confirm the developed vaccine's biological efficiency against XDR S. Typhi.Communicated by Ramaswamy H. Sarma.

Automatic Classification of Mass Shape and Margin on Mammography with Artificial Intelligence: Deep CNN Versus Radiomics

The purpose of this study is to test the feasibility for deep CNN-based artificial intelligence methods for automatic classification of the mass margin and shape, while radiomic feature-based machine learning methods were also implemented in this study as baseline and for comparison study. In this retrospective study, 596 patients with breast mass that underwent mammography from 4 hospitals were enrolled from January 2012 to October 2019. Margin and shape of each mass were annotated according to BI-RADS by 2 experienced radiologists. Deep CNN-based AI was implemented for the classification task based on Resnet50. Balanced sampler and CBAM were also used to improve the performance of the Deep CNNs. As comparison, image texture features were extracted and then dimensionality reduction methods (such as PCA, ICA) and classical classifiers (such as SVM, DT, KNN) were used for classification task. Based on Python programming software, accuracy (ACC) was used to evaluate the performance of the model, and the model with the highest ACC value was selected. Deep CNN based on Resnet50 with balanced sampler and CBAM achieved the best performance for both margin and shape classification, with ACC of 0.838 and 0.874, respectively. For the radiomics-based machine learning, the best performance for margin was achieved as 0.676 by the combination of FA + RF, while the best performance for shape was 0.802 by the combination of PCA + MLP. The feasibility for automatic classification with coarse labeling of the mass shape and margin was testified with the deep CNN-based AI methods, while radiomic feature-based machine learning methods achieved inferior classification results.

E, Du Y, Liu F, Cui W, Yang S, Chen X, Han J, Xie Q, Feng Y, Liu W, Tang P, Zhang J, Zheng J, Chen D, Yao X, Ren T, Li Y, Li Y, Wu L, Song Q, Yang M, Zhang J, Liu Y, Guo S, Yan K, Shen X, Lei D, Zhang Y, Li Y, Dong Y, Tang S. Epidemiology and Association Rules Analysis for Pulmonary Tuberculosis Cases with Extrapulmonary Tuberculosis from Age and Gender Perspective: A Large-Scale Retrospective Multicenter Observational Study in China

Background

Tuberculosis (TB), a multisystemic disease with protean presentation, remains a major global health problem. Although concurrent pulmonary tuberculosis (PTB) and extrapulmonary tuberculosis (EPTB) cases are commonly observed clinically, knowledge regarding concurrent PTB-EPTB is limited. Here, a large-scale multicenter observational study conducted in China aimed to study the epidemiology of concurrent PTB-EPTB cases by diagnostically defining TB types and then implementing association rules analysis.

Methods

The retrospective study was conducted at 21 hospitals in 15 provinces in China and included all inpatients with confirmed TB diagnoses admitted from Jan 2011 to Dec 2017. Association rules analysis was conducted for cases with concurrent PTB and various types of EPTB using the Apriori algorithm.

Results

Evaluation of 438,979TB inpatients indicated PTB was the most commonly diagnosed (82.05%) followed by tuberculous pleurisy (23.62%). Concurrent PTB-EPTB was found in 129,422 cases (29.48%) of which tuberculous pleurisy was the most common concurrent EPTB type observed. The multivariable logistic regression models demonstrated that odds ratios of concurrent PTB-EPTB cases varied by gender and age group. For PTB cases with concurrent EPTB, the strongest association was found between PTB and concurrent bronchial tuberculosis (lift = 1.09). For EPTB cases with concurrent PTB, the strongest association was found between pharyngeal/laryngeal tuberculosis and concurrent PTB (lift = 1.11). Confidence and lift values of concurrent PTB-EPTB cases varied with gender and age.

Conclusions

Numerous concurrent PTB-EPTB case types were observed, with confidence and lift values varying with gender and age. Clinicians should screen for concurrent PTB-EPTB in order to improve treatment outcomes.

Misclassified: identification of zoonotic transition biomarker candidates for influenza A viruses using deep neural network

Introduction: Zoonotic transition of Influenza A viruses is the cause of epidemics with high rates of morbidity and mortality. Predicting which viral strains are likely to transition from their genetic sequence could help in the prevention and response against these zoonotic strains. We hypothesized that features predictive of viral hosts could be leveraged to identify biomarkers of zoonotic viral transition. Methods: We trained deep learning models to predict viral hosts based on the virus mRNA or protein sequences. Our multi-host dataset contained 848,630 unique nucleotide sequences obtained from the NCBI Influenza Virus and Influenza Research Databases. Each sequence, representing one gene from one viral strain, was classified into one of the three host categories: Avian, Human, and Swine. Trained models were analyzed using various neural network interpretation methods to identify interesting candidates for zoonotic transition biomarkers. Results: Using mRNA sequences as input led to higher prediction accuracies than amino acids, suggesting that the codon sequence contains information relevant to viral hosts that is lost during protein translation. UMAP visualization of the latent space of our classifiers showed that viral sequences clustered according to their host of origin. Interestingly, sequences from pandemic zoonotic viral strains localized at the margins between hosts, while zoonotic sequences incapable of Human-to-Human transmission localized with non-zoonotic viruses from the same host. In addition, host prediction for pandemic zoonotic sequences had low prediction accuracy, which was not the case for the other zoonotic strains. This supports our hypothesis that ambiguously predicted viral sequences bear features associated with cross-species infectivity. Finally, we compared misclassified sequences to well-classified ones to extract interesting candidates for zoonotic transition biomarkers. While features varied significantly between pairs of species and viral genes, several codons were conserved in Swine-to-Human and Avian-to-Human misclassified sequences, and in particular in the NA, HA, and NP genes, suggesting their importance for zoonosis in Humans. Discussion: Analysis of viral sequences using neural network interpretation approaches revealed important genetic differences between zoonotic viruses with pandemic potential, compared to non-zoonotic viral strains or zoonotic viruses incapable of Human-to-Human transmission.

A First Computational Frame for Recognizing Heparin-Binding Protein

Heparin-binding protein (HBP) is a cationic antibacterial protein derived from multinuclear neutrophils and an important biomarker of infectious diseases. The correct identification of HBP is of great significance to the study of infectious diseases. This work provides the first HBP recognition framework based on machine learning to accurately identify HBP. By using four sequence descriptors, HBP and non-HBP samples were represented by discrete numbers. By inputting these features into a support vector machine (SVM) and random forest (RF) algorithm and comparing the prediction performances of these methods on training data and independent test data, it is found that the SVM-based classifier has the greatest potential to identify HBP. The model could produce an auROC of 0.981 ± 0.028 on training data using 10-fold cross-validation and an overall accuracy of 95.0% on independent test data. As the first model for HBP recognition, it will provide some help for infectious diseases and stimulate further research in related fields.

FastKnock: An efficient next-generation approach to identify all knockout strategies for strain optimization

Overproduction of desired native or nonnative biochemical(s) in (micro)organisms can be achieved through metabolic engineering. Appropriate rewiring of cell metabolism is performed making rational changes such as insertion, up-/down-regulation and knockout of genes and consequently metabolic reactions. Finding appropriate targets (including proper sets of reactions to be knocked out) for metabolic engineering to design optimal production strains has been the goal of a number of computational algorithms. We developed FastKnock, an efficient next-generation algorithm for identifying all possible knockout strategies for the growth-coupled overproduction of biochemical(s) of interest. We achieve this by developing a special depth-first traversal algorithm that allows us to prune the search space significantly. This leads to a drastic reduction in execution time. We evaluate the performance of the FastKnock algorithm using three Escherichia coli genome-scale metabolic models in different conditions (minimal and rich mediums) for the overproduction of a number of desired metabolites. FastKnock efficiently prunes the search space to less than 0.2% for quadruple and 0.02% for quintuple-reaction knockouts. Compared to the classic approaches such as OptKnock and the state-of-the-art techniques such as MCSEnumerator methods, FastKnock found many more useful and important practical solutions. The availability of all the solutions provides the opportunity to further characterize and select the most appropriate intervention strategy based on any desired evaluation index. Our implementation of the FastKnock method in Python is publicly available at https://github.com/leilahsn/FastKnock.

Transferring From Textual Entailment to Biomedical Named Entity Recognition

Biomedical Named Entity Recognition (BioNER) aims at identifying biomedical entities such as genes, proteins, diseases, and chemical compounds in the given textual data. However, due to the issues of ethics, privacy, and high specialization of biomedical data, BioNER suffers from the more severe problem of lacking in quality labeled data than the general domain especially for the token-level. Facing the extremely limited labeled biomedical data, this work studies the problem of gazetteer-based BioNER, which aims at building a BioNER system from scratch. It needs to identify the entities in the given sentences when we have zero token-level annotations for training. Previous works usually use sequential labeling models to solve the NER or BioNER task and obtain weakly labeled data from gazetteers when we don't have full annotations. However, these labeled data are quite noisy since we need the labels for each token and the entity coverage of the gazetteers is limited. Here we propose to formulate the BioNER task as a Textual Entailment problem and solve the task via Textual Entailment with Dynamic Contrastive learning (TEDC). TEDC not only alleviates the noisy labeling issue, but also transfers the knowledge from pre-trained textual entailment models. Additionally, the dynamic contrastive learning framework contrasts the entities and non-entities in the same sentence and improves the model's discrimination ability. Experiments on two real-world biomedical datasets show that TEDC can achieve state-of-the-art performance for gazetteer-based BioNER.

An Exploration of the Coherent Effects between METTL3 and NDUFA10 on Alzheimer's Disease

Alzheimer's disease (AD) is a neurodegenerative disorder characterized primarily by a decline in cognitive function. However, the etiopathogenesis of AD is unclear. N6-methyladenosine (m6A) is abundant in the brain, and it is interesting to explore the relationship between m6A and AD causes. In this paper, the gene expression of METTL3 and NDUFA10 were found to correlate with the Mini-mental State Examination (MMSE), which is a clinical indicator of the degree of dementia. METTL3 is involved in post-transcriptional methylation and the formation of m6A. NDUFA10 encodes the protein with NADH dehydrogenase activity and oxidoreductase activity in the mitochondrial electron transport chain. The following three characteristics were observed in this paper: 1. The lower the expression level of NDUFA10, the smaller the MMSE, and the higher the degree of dementia. 2. If the expression level of METTL3 dropped below its threshold, the patient would have a risk of AD with a probability close to 100%, suggesting a basic necessity for m6A to protect mRNA. 3. The lower the expression levels of both METTL3 and NDUFA10, the more likely the patient would suffer from AD, implying the coherence between METTL3 and NDUFA10. Regarding the above discovery, the following hypothesis is presented: METTL3 expression level is downregulated, then the m6A modification level of NDUFA10 mRNA is also decreased, thereby reducing the expression level of NDUFA10-encoded protein. Furthermore, the abnormal expression of NDUFA10 contributes to the assembly disorder of mitochondrial complex I and affects the process of the electron respiratory chain, with the consequent development of AD. In addition, to confirm the above conclusions, the AI Ant Colony Algorithm was improved to be more suitable for discovering the characteristics of AD data, and the SVM diagnostic model was applied to mine the coherent effects on AD between METTL3 and NDUFA10. In conclusion, our findings suggest that dysregulated m6A leads to altered expression of its target genes, thereby affecting AD's development.

Reservoir based spiking models for univariate Time Series Classification

A variety of advanced machine learning and deep learning algorithms achieve state-of-the-art performance on various temporal processing tasks. However, these methods are heavily energy inefficient-they run mainly on the power hungry CPUs and GPUs. Computing with Spiking Networks, on the other hand, has shown to be energy efficient on specialized neuromorphic hardware, e.g., Loihi, TrueNorth, SpiNNaker, etc. In this work, we present two architectures of spiking models, inspired from the theory of Reservoir Computing and Legendre Memory Units, for the Time Series Classification (TSC) task. Our first spiking architecture is closer to the general Reservoir Computing architecture and we successfully deploy it on Loihi; the second spiking architecture differs from the first by the inclusion of non-linearity in the readout layer. Our second model (trained with Surrogate Gradient Descent method) shows that non-linear decoding of the linearly extracted temporal features through spiking neurons not only achieves promising results, but also offers low computation-overhead by significantly reducing the number of neurons compared to the popular LSM based models-more than 40x reduction with respect to the recent spiking model we compare with. We experiment on five TSC datasets and achieve new SoTA spiking results (-as much as 28.607% accuracy improvement on one of the datasets), thereby showing the potential of our models to address the TSC tasks in a green energy-efficient manner. In addition, we also do energy profiling and comparison on Loihi and CPU to support our claims.

Main Factors Shaping Amino Acid Usage Across Evolution

The standard genetic code determines that in most species, including viruses, there are 20 amino acids that are coded by 61 codons, while the other three codons are stop triplets. Considering the whole proteome each species features its own amino acid frequencies, given the slow rate of change, closely related species display similar GC content and amino acids usage. In contrast, distantly related species display different amino acid frequencies. Furthermore, within certain multicellular species, as mammals, intragenomic differences in the usage of amino acids are evident. In this communication, we shall summarize some of the most prominent and well-established factors that determine the differences found in the amino acid usage, both across evolution and intragenomically.

Looking beyond Virus Detection in RNA Sequencing Data: Lessons Learned from a Community-Based Effort to Detect Cellular Plant Pathogens and Pests

High-throughput sequencing (HTS), more specifically RNA sequencing of plant tissues, has become an indispensable tool for plant virologists to detect and identify plant viruses. During the data analysis step, plant virologists typically compare the obtained sequences to reference virus databases. In this way, they are neglecting sequences without homologies to viruses, which usually represent the majority of sequencing reads. We hypothesized that traces of other pathogens might be detected in this unused sequence data. In the present study, our goal was to investigate whether total RNA-seq data, as generated for plant virus detection, is also suitable for the detection of other plant pathogens and pests. As proof of concept, we first analyzed RNA-seq datasets of plant materials with confirmed infections by cellular pathogens in order to check whether these non-viral pathogens could be easily detected in the data. Next, we set up a community effort to re-analyze existing Illumina RNA-seq datasets used for virus detection to check for the potential presence of non-viral pathogens or pests. In total, 101 datasets from 15 participants derived from 51 different plant species were re-analyzed, of which 37 were selected for subsequent in-depth analyses. In 29 of the 37 selected samples (78%), we found convincing traces of non-viral plant pathogens or pests. The organisms most frequently detected in this way were fungi (15/37 datasets), followed by insects (13/37) and mites (9/37). The presence of some of the detected pathogens was confirmed by independent (q)PCRs analyses. After communicating the results, 6 out of the 15 participants indicated that they were unaware of the possible presence of these pathogens in their sample(s). All participants indicated that they would broaden the scope of their bioinformatic analyses in future studies and thus check for the presence of non-viral pathogens. In conclusion, we show that it is possible to detect non-viral pathogens or pests from total RNA-seq datasets, in this case primarily fungi, insects, and mites. With this study, we hope to raise awareness among plant virologists that their data might be useful for fellow plant pathologists in other disciplines (mycology, entomology, bacteriology) as well.

Metal Body Burden as Cardiovascular Risk Factor in Adults with Metabolic Syndrome and Overweight-Obesity Analysed with an Artificial Neural Network: The Role of Hair Mineralograms

In determining the so-called "body burden", hair has been widely accepted for assessing toxic element exposure. However, its role in assessing essential elements is controversial. This study investigates the possible relationship between hair minerals, metabolic syndrome (MetS) and cardiovascular (CV) risk in non-occupationally exposed subjects with overweight-obesity. Ninety-five voluntary participants (aged 51 ± 12) were recruited in Northern Italy. Hair samples were collected and analysed via inductively coupled plasma mass spectrometry; the total toxicity index (TI) was calculated as well. To evaluate cardiovascular risk factors in the presence or absence of MetS, the following factors were considered via the innovative artificial neural network (ANN) method Auto-CM: hair mineralograms (31 elements) and 25 variables including blood pressure, anthropometric parameters, insulin resistance and biochemical serum markers assessing inflammation. The Framingham risk score, fatty liver index (FLI), visceral adiposity index and CV risk scores were also taken into consideration. As shown by the semantic map, which was subsequently confirmed by an activation and competition system (ACS), obesity parameters are strictly associated with CV risk factors, TI and inflammation; meanwhile, the single mineral elements seem to be unimportant. Data obtained via ANN demonstrate that MetS may be at least partly mediated by altered mineral levels also in the presence of obesity and that waist circumference is a crucial point to be monitored rather than BMI alone. Furthermore, the mineral body burden is one of the important factors for CV risk.

Machine learning and disease prediction in obstetrics

Machine learning technologies and translation of artificial intelligence tools to enhance the patient experience are changing obstetric and maternity care. An increasing number of predictive tools have been developed with data sourced from electronic health records, diagnostic imaging and digital devices. In this review, we explore the latest tools of machine learning, the algorithms to establish prediction models and the challenges to assess fetal well-being, predict and diagnose obstetric diseases such as gestational diabetes, pre-eclampsia, preterm birth and fetal growth restriction. We discuss the rapid growth of machine learning approaches and intelligent tools for automated diagnostic imaging of fetal anomalies and to asses fetoplacental and cervix function using ultrasound and magnetic resonance imaging. In prenatal diagnosis, we discuss intelligent tools for magnetic resonance imaging sequencing of the fetus, placenta and cervix to reduce the risk of preterm birth. Finally, the use of machine learning to improve safety standards in intrapartum care and early detection of complications will be discussed. The demand for technologies to enhance diagnosis and treatment in obstetrics and maternity should improve frameworks for patient safety and enhance clinical practice.

The use of artificial intelligence for automating or semi-automating biomedical literature analyses: a scoping review

OBJECTIVE

Evidence-based medicine (EBM) is a decision-making process based on the conscious and judicious use of the best available scientific evidence. However, the exponential increase in the amount of information currently available likely exceeds the capacity of human-only analysis. In this context, artificial intelligence (AI) and its branches such as machine learning (ML) can be used to facilitate human efforts in analyzing the literature to foster EBM. The present scoping review aimed to examine the use of AI in the automation of biomedical literature survey and analysis with a view to establishing the state-of-the-art and identifying knowledge gaps.

MATERIALS AND METHODS

Comprehensive searches of the main databases were performed for articles published up to June 2022 and studies were selected according to inclusion and exclusion criteria. Data were extracted from the included articles and the findings categorized.

RESULTS

The total number of records retrieved from the databases was 12,145, of which 273 were included in the review. Classification of the studies according to the use of AI in evaluating the biomedical literature revealed three main application groups, namely assembly of scientific evidence (n=127; 47%), mining the biomedical literature (n=112; 41%) and quality analysis (n=34; 12%). Most studies addressed the preparation of systematic reviews, while articles focusing on the development of guidelines and evidence synthesis were the least frequent. The biggest knowledge gap was identified within the quality analysis group, particularly regarding methods and tools that assess the strength of recommendation and consistency of evidence.

CONCLUSION

Our review shows that, despite significant progress in the automation of biomedical literature surveys and analyses in recent years, intense research is needed to fill knowledge gaps on more difficult aspects of ML, deep learning and natural language processing, and to consolidate the use of automation by end-users (biomedical researchers and healthcare professionals).

Machine learning-based glycolysis-associated molecular classification reveals differences in prognosis, TME, and immunotherapy for colorectal cancer patients

Background

Aerobic glycolysis is a process that metabolizes glucose under aerobic conditions, finally producing pyruvate, lactic acid, and ATP for tumor cells. Nevertheless, the overall significance of glycolysis-related genes in colorectal cancer and how they affect the immune microenvironment have not been investigated.

Methods

By combining the transcriptome and single-cell analysis, we summarize the various expression patterns of glycolysis-related genes in colorectal cancer. Three glycolysis-associated clusters (GAC) were identified with distinct clinical, genomic, and tumor microenvironment (TME). By mapping GAC to single-cell RNA sequencing analysis (scRNA-seq), we next discovered that the immune infiltration profile of GACs was similar to that of bulk RNA sequencing analysis (bulk RNA-seq). In order to determine the kind of GAC for each sample, we developed the GAC predictor using markers of single cells and GACs that were most pertinent to clinical prognostic indications. Additionally, potential drugs for each GAC were discovered using different algorithms.

Results

GAC1 was comparable to the immune-desert type, with a low mutation probability and a relatively general prognosis; GAC2 was more likely to be immune-inflamed/excluded, with more immunosuppressive cells and stromal components, which also carried the risk of the poorest prognosis; Similar to the immune-activated type, GAC3 had a high mutation rate, more active immune cells, and excellent therapeutic potential.

Conclusion

In conclusion, we combined transcriptome and single-cell data to identify new molecular subtypes using glycolysis-related genes in colorectal cancer based on machine-learning methods, which provided therapeutic direction for colorectal patients.

AI-Based Decision Support System for Traumatic Brain Injury: A Survey

Traumatic brain injury (TBI) is one of the major causes of disability and mortality worldwide. Rapid and precise clinical assessment and decision-making are essential to improve the outcome and the resulting complications. Due to the size and complexity of the data analyzed in TBI cases, computer-aided data processing, analysis, and decision support systems could play an important role. However, developing such systems is challenging due to the heterogeneity of symptoms, varying data quality caused by different spatio-temporal resolutions, and the inherent noise associated with image and signal acquisition. The purpose of this article is to review current advances in developing artificial intelligence-based decision support systems for the diagnosis, severity assessment, and long-term prognosis of TBI complications.

Public opinion changing patterns under the double-hazard scenario of natural disaster and public health event

In the context of the COVID-19 epidemic, a "double-hazard scenario" consisting of a natural disaster and a public health event occurring simultaneously is likely to arise. Focusing on this double-hazard scenario, this study developed a new opinion dynamics model that verifies the effect of opinion dynamic in practical applications and extends the realistic meaning of the logic matrix. The new model can be used to quickly identify changing trends in public opinion about two co-occurring public safety events in China, helping the government to better anticipate and respond to these real double-hazard scenarios. The new model was tested with three real double-hazard scenarios involving natural disasters and public health events in China and the simulation results were analyzed. Using visualization and Pearson correlation coefficients to analyze more than a million items of network-wide public opinion data, the new model was found to show a good fit with reality. The study finally found that in China, public attention to both natural hazards and public health events was greater when these public safety events co-occurred (double-hazard scenario) than when they occurred separately (single-hazard scenarios). These results verify the coupling phenomenon of different disasters in a multi-hazard scenario at the information level for the first time, which is greatly meaningful for multi-hazard research.

Recent Advances in Machine Learning-Based Models for Prediction of Antiviral Peptides

Viruses have killed and infected millions of people across the world. It causes several chronic diseases like COVID-19, HIV, and hepatitis. To cope with such diseases and virus infections, antiviral peptides (AVPs) have been applied in the design of drugs. Keeping in view the significant role in pharmaceutical industry and other research fields, identification of AVPs is highly indispensable. In this connection, experimental and computational methods were proposed to identify AVPs. However, more accurate predictors for boosting AVPs identification are highly desirable. This work presents a thorough study and reports the available predictors of AVPs. We explained applied datasets, feature representation approaches, classification algorithms, and evaluation parameters of performance. In this study, the limitations of the existing studies and the best methods were emphasized. Provided the pros and cons of the applied classifiers. The future insights demonstrate efficient feature encoding approaches, best feature optimization schemes, and effective classification techniques that can improve the performance of novel method for accurate prediction of AVPs.

Gene regulatory networks with binary weights

An evolutionary computation framework to learn binary threshold networks is presented. Inspired by the recent trend of binary neural networks, where weights and activation thresholds are represented using 1 and -1 such that they can be stored in 1-bit instead of full precision, we explore this approach for gene regulatory network modeling. We test our method by inferring binary threshold networks of two regulatory network models: Quorum sensing systems in bacterium Paraburkholderia phytofirmans PsJN and the fission yeast cell-cycle. We considered differential evolution and particle swarm optimization for the simulations. Results for weights having only 1 and -1 values, and different activation thresholds are presented. Full binary threshold networks were found with minimum error (2 bits), whereas when the binary restriction is relaxed for the activation thresholds, networks with 0 bit error were found. .