Disorder-specific versus transdiagnostic cognitive mechanisms in anxiety and depression: Machine-learning-based prediction of symptom severity

INTRODUCTION

Psychiatric evaluation of anxiety and depression is currently based on self-reported symptoms and their classification into discrete disorders. Yet the substantial overlap between these disorders as well as their within-disorder heterogeneity may contribute to the mediocre success rates of treatments. The proposed research examines a new framework for diagnosis that is based on alterations in underlying cognitive mechanisms. In line with the Research Domain Criteria (RDoC) approach, the current study directly compares disorder-specific and transdiagnostic cognitive patterns in predicting the severity of anxiety and depression symptoms.

METHODS

The sample included 237 individuals exhibiting differing levels of anxiety and depression symptoms, as measured by the STAI-T and BDI-II. Random Forest regressors were used to analyze their performance on a battery of six computerized cognitive-behavioral tests targeting selective and spatial attention, expectancy, interpretation, memory, and cognitive control biases.

RESULTS

Unique anxiety-specific biases were found, as well as shared anxious-depressed bias patterns. These cognitive biases exhibited relatively high fitting rates when predicting symptom severity (questionnaire scores common range 0-60, MAE = 6.03, RMSE = 7.53). Interpretation and expectancy biases exhibited the highest association with symptoms, above all other individual biases.

LIMITATIONS

Although internal validation methods were applied, models may suffer from potential overfitting due to sample size limitations.

CONCLUSION

In the context of the ongoing dispute regarding symptom-centered versus transdiagnostic approaches, the current study provides a unique comparison of these two views, yielding a novel intermediate approach. The results support the use of mechanism-based dimensional diagnosis for adding precision and objectivity to future psychiatric evaluations.

Discovery and validation of molecular patterns and immune characteristics in the peripheral blood of ischemic stroke patients

Background

Stroke is a disease with high morbidity, disability, and mortality. Immune factors play a crucial role in the occurrence of ischemic stroke (IS), but their exact mechanism is not clear. This study aims to identify possible immunological mechanisms by recognizing immune-related biomarkers and evaluating the infiltration pattern of immune cells.

Methods

We downloaded datasets of IS patients from GEO, applied R language to discover differentially expressed genes, and elucidated their biological functions using GO, KEGG analysis, and GSEA analysis. The hub genes were then obtained using two machine learning algorithms (least absolute shrinkage and selection operator (LASSO) and support vector machine-recursive feature elimination (SVM-RFE)) and the immune cell infiltration pattern was revealed by CIBERSORT. Gene-drug target networks and mRNA-miRNA-lncRNA regulatory networks were constructed using Cytoscape. Finally, we used RT-qPCR to validate the hub genes and applied logistic regression methods to build diagnostic models validated with ROC curves.

Results

We screened 188 differentially expressed genes whose functional analysis was enriched to multiple immune-related pathways. Six hub genes (ANTXR2, BAZ2B, C5AR1, PDK4, PPIH, and STK3) were identified using LASSO and SVM-RFE. ANTXR2, BAZ2B, C5AR1, PDK4, and STK3 were positively correlated with neutrophils and gamma delta T cells, and negatively correlated with T follicular helper cells and CD8, while PPIH showed the exact opposite trend. Immune infiltration indicated increased activity of monocytes, macrophages M0, neutrophils, and mast cells, and decreased infiltration of T follicular helper cells and CD8 in the IS group. The ceRNA network consisted of 306 miRNA-mRNA interacting pairs and 285 miRNA-lncRNA interacting pairs. RT-qPCR results indicated that the expression levels of BAZ2B, C5AR1, PDK4, and STK3 were significantly increased in patients with IS. Finally, we developed a diagnostic model based on these four genes. The AUC value of the model was verified to be 0.999 in the training set and 0.940 in the validation set.

Conclusion

Our research explored the immune-related gene expression modules and provided a specific basis for further study of immunomodulatory therapy of IS.

Accurate and robust segmentation of cerebral vasculature on four-dimensional arterial spin labeling magnetic resonance angiography using machine-learning approach

Segmentation of cerebral vasculature on MR vascular images is of great significance for clinical application and research. However, the existing cerebrovascular segmentation approaches are limited due to insufficient image contrast and complicated algorithms. This study aims to explore the potential of the emerging four-dimensional arterial spin labeling magnetic resonance angiography (4D ASL-MRA) technique for fast and accurate cerebrovascular segmentation with a simple machine-learning approach. Nine temporal features were extracted from the intensity-time signal of each voxel, and eight spatial features from the neighboring voxels. Then, the unsupervised outlier detection algorithm, i.e. Isolation Forest, is used for segmentation of the vascular voxels based on the extracted features. The total length of the centerlines of the intracranial arterial vasculature, the dice similarity coefficient (DSC), and the average Hausdorff Distance (AVGHD) on the cross-sections of small- to large-sized vessels were calculated to evaluate the performance of the segmentation approach on 4D ASL-MRA of 18 subjects. Experiments show that the temporal information on 4D ASL-MRA can largely improve the segmentation performance. In addition, the proposed segmentation approach outperforms the traditional methods that were performed on the 3D image (i.e. the temporal average intensity projection of 4D ASL-MRA) and the previously proposed frame-wise approach. In conclusion, this study demonstrates that accurate and robust segmentation of cerebral vasculature is achievable on 4D ASL-MRA by using a simple machine-learning approach with appropriate features.

Performance and robustness of small molecule retention time prediction with molecular graph neural networks in industrial drug discovery campaigns

This study explores how machine-learning can be used to predict chromatographic retention times (RT) for the analysis of small molecules, with the objective of identifying a machine-learning framework with the robustness required to support a chemical synthesis production platform. We used internally generated data from high-throughput parallel synthesis in context of pharmaceutical drug discovery projects. We tested machine-learning models from the following frameworks: XGBoost, ChemProp, and DeepChem, using a dataset of 7552 small molecules. Our findings show that two specific models, AttentiveFP and ChemProp, performed better than XGBoost and a regular neural network in predicting RT accurately. We also assessed how well these models performed over time and found that molecular graph neural networks consistently gave accurate predictions for new chemical series. In addition, when we applied ChemProp on the publicly available METLIN SMRT dataset, it performed impressively with an average error of 38.70 s. These results highlight the efficacy of molecular graph neural networks, especially ChemProp, in diverse RT prediction scenarios, thereby enhancing the efficiency of chromatographic analysis.

An update on antibacterial AlkylGuanidino Ureas: Design of new derivatives, synergism with colistin and data analysis of the whole library

Antimicrobial resistance (AMR) represents one of the most challenging global Public Health issues, with an alarmingly increasing rate of attributable mortality. This scenario highlights the urgent need for innovative medicinal strategies showing activity on resistant isolates (especially, carbapenem-resistant Gram-negative bacteria, methicillin-resistant S. aureus, and vancomycin-resistant enterococci) yielding new approaches for the treatment of bacterial infections. We previously reported AlkylGuanidino Ureas (AGUs) with broad-spectrum antibacterial activity and a putative membrane-based mechanism of action. Herein, new tetra- and mono-guanidino derivatives were designed and synthesized to expand the structure-activity relationships (SARs) and, thereby, tested on the same panel of Gram-positive and Gram-negative bacteria. The membrane-active mechanism of selected compounds was then investigated through molecular dynamics (MD) on simulated bacterial membranes. In the end, the newly synthesized series, along with the whole library of compounds (more than 70) developed in the last decade, was tested in combination with subinhibitory concentrations of the last resort antibiotic colistin to assess putative synergistic or additive effects. Moreover, all the AGUs were subjected to cheminformatic and machine learning analyses to gain a deeper knowledge of the key features required for bioactivity.

Potential distribution of Biscogniauxia mediterranea and Obolarina persica causal agents of oak charcoal disease in Iran's Zagros forests

In Iran, native oak species are under threat from episodes of Charcoal Disease, a decline syndrome driven by abiotic stressors (e.g. drought, elevated temperature) and biotic components, Biscogniauxia mediterranea (De Not.) Kuntze and Obolarina persica (M. Mirabolfathy). The outbreak is still ongoing and the country's largest ever recorded. Still, the factors driving its' epidemiology in time and space are poorly known and such knowledge is urgently needed to develop strategies to counteract the adverse effects. In this study, we developed a generic framework based on experimental, machine-learning algorithms and spatial analyses for landscape-level prediction of oak charcoal disease outbreaks. Extensive field surveys were conducted during 2013-2015 in eight provinces (more than 50 unique counties) in the Zagros ecoregion. Pathogenic fungi were isolated and characterized through morphological and molecular approaches, and their pathogenicity was assessed under controlled water stress regimes in the greenhouse. Further, we evaluated a set of 29 bioclimatic, environmental, and host layers in modeling for disease incidence data using four well-known machine learning algorithms including the Generalized Linear Model, Gradient Boosting Model, Random Forest model (RF), and Multivariate Adaptive Regression Splines implemented in MaxEnt software. Model validation statistics [Area Under the Curve (AUC), True Skill Statistics (TSS)], and Kappa index were used to evaluate the accuracy of each model. Models with a TSS above 0.65 were used to prepare an ensemble model. The results showed that among the different climate variables, precipitation and temperature (Bio18, Bio7, Bio8, and bio9) in the case of O. persica and similarly, gsl (growing season length TREELIM, highlighting the warming climate and the endophytic/pathogenic nature of the fungus) and precipitation in case of B. mediterranea are the most important influencing variables in disease modeling, while near-surface wind speed (sfcwind) is the least important variant. The RF algorithm generates the most robust predictions (ROC of 0.95; TSS of 0.77 and 0.79 for MP and OP, respectively). Theoretical analysis shows that the ensemble model (ROC of 0.95 and 0.96; TSS = 0.79 and 0.81 for MP and OP, respectively), can efficiently be used in the prediction of the charcoal disease spatiotemporal distribution. The oak mortality varied ranging from 2 to 14%. Wood-boring beetles association with diseased trees was determined at 20%. Results showed that water deficiency is a crucial component of the oak decline phenomenon in Iran. The Northern Zagros forests (Ilam, Lorestan, and Kermanshah provinces) along with the southern Zagros forests (Fars and Kohgilouyeh va-Boyer Ahmad provinces) among others are the most endangered areas of potential future pandemics of charcoal disease. Our findings will significantly improve our understanding of the current situation of the disease to pave the way against pathogenic agents in Iran.

BrainAGE, brain health, and mental disorders: A systematic review

The imaging-based method of brainAGE aims to characterize an individual's vulnerability to age-related brain changes. The present study systematically reviewed brainAGE findings in neuropsychiatric conditions and discussed the potential of brainAGE as a marker for biological age. A systematic PubMed search (from inception to March 6th, 2023) identified 273 articles. The 30 included studies compared brainAGE between neuropsychiatric and healthy groups (n≥50). We presented results qualitatively and adapted a bias risk assessment questionnaire. The imaging modalities, design, and input features varied considerably between studies. While the studies found higher brainAGE in neuropsychiatric conditions (11 mild cognitive impairment/ dementia, 11 schizophrenia spectrum/ other psychotic and bipolar disorder, six depression/ anxiety, two multiple groups), the associations with clinical characteristics were mixed. While brainAGE is sensitive to group differences, limitations include the lack of diverse training samples, multi-modal studies, and external validation. Only a few studies obtained longitudinal data, and all have used algorithms built solely to predict chronological age. These limitations impede the validity of brainAGE as a biological age marker.

Prospective virtual screening combined with bio-molecular simulation enabled identification of new inhibitors for the KRAS drug target

Lung cancer is a disease with a high mortality rate and it is the number one cause of cancer death globally. Approximately 12-14% of non-small cell lung cancers are caused by mutations in KRASG12C. The KRASG12C is one of the most prevalent mutants in lung cancer patients. KRAS was first considered undruggable. The sotorasib and adagrasib are the recently approved drugs that selectively target KRASG12C, and offer new treatment approaches to enhance patient outcomes however drug resistance frequently arises. Drug development is a challenging, expensive, and time-consuming process. Recently, machine-learning-based virtual screening are used for the development of new drugs. In this study, we performed machine-learning-based virtual screening followed by molecular docking, all atoms molecular dynamics simulation, and binding energy calculations for the identifications of new inhibitors against the KRASG12C mutant. In this study, four machine learning models including, random forest, k-nearest neighbors, Gaussian naïve Bayes, and support vector machine were used. By using an external dataset and 5-fold cross-validation, the developed models were validated. Among all the models the performance of the random forest (RF) model was best on the train/test dataset and external dataset. The random forest model was further used for the virtual screening of the ZINC15 database, in-house database, Pakistani phytochemicals, and South African Natural Products database. A total of 100 ns MD simulation was performed for the four best docking score complexes as well as the standard compound in complex with KRASG12C. Furthermore, the top four hits revealed greater stability and greater binding affinities for KRASG12C compared to the standard drug. These new hits have the potential to inhibit KRASG12C and may help to prevent KRAS-associated lung cancer. All the datasets used in this study can be freely available at ( https://github.com/Amar-Ajmal/Datasets-for-KRAS ).

Identification of a novel ADCC-related gene signature for predicting the prognosis and therapy response in lung adenocarcinoma

BACKGROUND

Previous studies have largely neglected the role of ADCC in LUAD, and no study has systematically compiled ADCC-associated genes to create prognostic signatures.

METHODS

In this study, 1564 LUAD patients, 2057 NSCLC patients, and more than 5000 patients with various cancer types from diverse cohorts were included. R package ConsensusClusterPlus was utilized to classify patients into different subtypes. A number of machine-learning algorithms were used to construct the ADCCRS. GSVA and ClusterProfiler were used for enrichment analyses, and IOBR was used to quantify immune cell infiltration level. GISTIC2.0 and maftools were used to analyze the CNV and SNV data. The Oncopredict package was used to predict drug information based on the GDSC1. Three immunotherapy cohorts were used to evaluate patient response to immunotherapy. The Seurat package was used to process single-cell data, the AUCell package was used to calculate cells' geneset activity scores, and the Scissor algorithm was used to identify ADCCRS-associated cells.

RESULTS

Through unsupervised clustering, two distinct subtypes of LUAD were identified, each exhibiting distinct clinical characteristics. The ADCCRS, consisted of 16 genes, was constructed by integrated machine-learning methods. The prognostic power of ADCCRS was validated in 28 independent datasets. Further, ADCCRS shows better predictive abilities than 102 previously published signatures in predicting LUAD patients' survival. A nomogram incorporating ADCCRS and clinical features was constructed, demonstrating high predictive performance. ADCCRS positively correlates with patients' gene mutation, and integrated analysis of bulk and single-cell transcriptome data revealed the association of ADCCRS with TME modulators. Cells representing high-ADCCRS phenotype exhibited more malignant features. LUAD patients with high ADCCRS levels exhibited sensitivity to chemotherapy and targeted therapy, while displaying resistance to immunotherapy. In pan-cancer analysis, ADCCRS still exhibited significant prognostic value and was found to be a risk factor for most cancer patients.

CONCLUSIONS

ADCCRS offers a critical prognostic insight for patients with LUAD, shedding light on the tumor microenvironment and forecasting treatment responsiveness.

Identification of 7 mitochondria-related genes as diagnostic biomarkers of MDD and their correlation with immune infiltration: New insights from bioinformatics analysis

BACKGROUND

Major depressive disorder (MDD) is one of the most prevalent and debilitating psychiatric disorders. It becomes more recognized that mitochondrial dysfunction contributes to the pathophysiology of depression. However, little research has systematically investigated the mitochondria-related biomarkers for MDD diagnosis. This study aimed to develop a novel diagnostic gene signature in MDD based on mitochondria-related genes.

METHOD

We identified the differentially expressed mitochondrial-related genes (DeMRGs) by combing the gene expression data of the GEO database with mitochondria-related gene lists obtained from the MitoCarta3.0 database. Next, three kinds of machine-learning algorithms were used to screen characteristic DeMRGs. Then, we constructed a multivariable diagnostic model based on these characteristic genes and evaluated the diagnostic ability of this model. Subsequently, the immune landscape of infiltrated immune cells between MDD patients and controls was evaluated by CIBERSORT. Using consensus clustering analysis, we divided MDD patients into different clusters based on the characteristic DeMRGs expression patterns. Finally, the variations in immune cell infiltration between different clusters, and the correlation between characteristic DeMRGs and immune cell infiltration were analyzed.

RESULTS

Seven characteristic genes, including PMPCB, MRPS28, LYRM2, MGST1, COX20, PTPMT1, and STX17, were identified from the 31 DeMRGs. Based on the seven characteristic genes, we successfully constructed a diagnostic model which had relatively good diagnostic performance and potential application in the clinical diagnosis of MDD. In addition, our results also imply an intimate and comprehensive association between the characteristic DeMRGs and immune infiltrating cells.

CONCLUSION

A novel mitochondria-related gene signature with a good diagnostic performance and a relationship with immune microenvironment were identified in major depressive disorder.

Cancer screening in hospitalized ischemic stroke patients: a multicenter study focused on multiparametric analysis to improve management of occult cancers

Background/aims

The reciprocal promotion of cancer and stroke occurs due to changes in shared risk factors, such as metabolic pathways and molecular targets, creating a "vicious cycle." Cancer plays a direct or indirect role in the pathogenesis of ischemic stroke (IS), along with the reactive medical approach used in the treatment and clinical management of IS patients, resulting in clinical challenges associated with occult cancer in these patients. The lack of reliable and simple tools hinders the effectiveness of the predictive, preventive, and personalized medicine (PPPM/3PM) approach. Therefore, we conducted a multicenter study that focused on multiparametric analysis to facilitate early diagnosis of occult cancer and personalized treatment for stroke associated with cancer.

Methods

Admission routine clinical examination indicators of IS patients were retrospectively collated from the electronic medical records. The training dataset comprised 136 IS patients with concurrent cancer, matched at a 1:1 ratio with a control group. The risk of occult cancer in IS patients was assessed through logistic regression and five alternative machine-learning models. Subsequently, select the model with the highest predictive efficacy to create a nomogram, which is a quantitative tool for predicting diagnosis in clinical practice. Internal validation employed a ten-fold cross-validation, while external validation involved 239 IS patients from six centers. Validation encompassed receiver operating characteristic (ROC) curves, calibration curves, decision curve analysis (DCA), and comparison with models from prior research.

Results

The ultimate prediction model was based on logistic regression and incorporated the following variables: regions of ischemic lesions, multiple vascular territories, hypertension, D-dimer, fibrinogen (FIB), and hemoglobin (Hb). The area under the ROC curve (AUC) for the nomogram was 0.871 in the training dataset and 0.834 in the external test dataset. Both calibration curves and DCA underscored the nomogram's strong performance.

Conclusions

The nomogram enables early occult cancer diagnosis in hospitalized IS patients and helps to accurately identify the cause of IS, while the promotion of IS stratification makes personalized treatment feasible. The online nomogram based on routine clinical examination indicators of IS patients offered a cost-effective platform for secondary care in the framework of PPPM.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-024-00354-8.

Assessing the robustness of a machine-learning model for early detection of pancreatic adenocarcinoma (PDA): evaluating resilience to variations in image acquisition and radiomics workflow using image perturbation methods

PURPOSE

To evaluate robustness of a radiomics-based support vector machine (SVM) model for detection of visually occult PDA on pre-diagnostic CTs by simulating common variations in image acquisition and radiomics workflow using image perturbation methods.

METHODS

Eighteen algorithmically generated-perturbations, which simulated variations in image noise levels (σ, 2σ, 3σ, 5σ), image rotation [both CT image and the corresponding pancreas segmentation mask by 45° and 90° in axial plane], voxel resampling (isotropic and anisotropic), gray-level discretization [bin width (BW) 32 and 64)], and pancreas segmentation (sequential erosions by 3, 4, 6, and 8 pixels and dilations by 3, 4, and 6 pixels from the boundary), were introduced to the original (unperturbed) test subset (n = 128; 45 pre-diagnostic CTs, 83 control CTs with normal pancreas). Radiomic features were extracted from pancreas masks of these additional test subsets, and the model's performance was compared vis-a-vis the unperturbed test subset.

RESULTS

The model correctly classified 43 out of 45 pre-diagnostic CTs and 75 out of 83 control CTs in the unperturbed test subset, achieving 92.2% accuracy and 0.98 AUC. Model's performance was unaffected by a three-fold increase in noise level except for sensitivity declining to 80% at 3σ (p = 0.02). Performance remained comparable vis-a-vis the unperturbed test subset despite variations in image rotation (p = 0.99), voxel resampling (p = 0.25-0.31), change in gray-level BW to 32 (p = 0.31-0.99), and erosions/dilations up to 4 pixels from the pancreas boundary (p = 0.12-0.34).

CONCLUSION

The model's high performance for detection of visually occult PDA was robust within a broad range of clinically relevant variations in image acquisition and radiomics workflow.

A Bayesian decision support system for optimizing pavement management programs

Over time, the pavement deteriorates due to traffic and the environment, resulting in poor riding quality and structural inadequacies. Evaluating pavement condition over time is thus a critical component of any pavement management system (PMS) to extend the service life of pavements. However, the uncertainty associated with the pavement deterioration process due to the heterogeneous nature of the pavement degradation factors makes the process difficult. The current work addresses this challenge of pavement management by developing an expert system framework based on Bayesian Belief Networks (BBN). This framework integrates data on existing road deterioration factors with knowledge gained from pavement experts to produce optimal decisions. The advantages of the BBN techniques lie in their ability to capture uncertainty, and probabilistically infer the values of variables in the domain, especially in the case of incomplete information where we only have data about some and not all variables. This has motivated the adoption of BBN in this study to optimize pavement maintenance decisions, on the basis of inferred road deterioration interpretations drawn from partial knowledge about road distress variables. This study presents the adoption of Bayesian methods to assist pavement maintenance engineers in determining the most successful and efficient maintenance and repair (M&R) tactics and the best time to apply them by means of a decision-support system. Data collected from 32 road sections in the United Arab Emirates in relation to road distress parameters (rutting, deflection, cracking, and international roughness index), as well as road characteristics, traffic, and environment data, has been used to demonstrate the applicability of the proposed decision-support tool.

Selecting, optimizing and externally validating a preexisting machine-learning regression algorithm for estimating waist circumference

Obesity, typically defined by the body mass index (BMI), has well known negative health effects. However, the BMI has serious deficiencies in predicting the adverse risks associated to obesity. Waist circumference (WC) is an alternative to define obesity and a better disease predictor according to the literature. However, old databases often lack this information, it is inaccurate (collected via self-report) or it is incomplete. Thus, this study accurately assesses WC using machine learning. The novel approaches are: 1) predictor variables (weight, height, age and sex) likely to appear in most data sets are used. 2) Publicly available data (including non-adults) and algorithms are used. 3) Systematic methods for data cleanup, model selection, hyperparameter optimization and external validation are performed. DATA ARE CLEANED: one variable per column, no special codes, missing values or outliers. Preexisting regression algorithms are gaged by cross-validation, using one data set. The hyperparameters of the best performing algorithm are optimized. The tuned algorithm is externally validated with other data sets by cross-validation. In spite of the limited number of features, the tuned algorithm outperforms prior WC approximations, using the same or similar predictor variables. The tuned algorithm enables using data where WC is not measured, is incomplete or is unreliable. A similar approach would be useful to estimate other variables of interest.

Development of the next-generation functional neuro-cognitive imaging protocol - Part 1: A 3D sliding-window convolutional neural net for automated brain parcellation

Functional MRI has emerged as a powerful tool to assess the severity of Post-concussion syndrome (PCS) and to provide guidance for neuro-cognitive therapists during treatment. The next-generation functional neuro-cognitive imaging protocol (fNCI2) has been developed to provide this assessment. This paper covers the first step in the analysis process, the development of a rapidly re-trainable, machine-learning, brain parcellation tool. The use of a sufficiently deep U-Net architecture encompassing a small (39 × 39 × 39 voxel input, 27 × 27 × 27 voxel output) sliding window to sample the entirety of the 3D image allows for the prediction of the entire image using only a single trained network. A large number of training, validating, and testing windows are thus generated from the 101 manually-labeled Mindboggle images, and full-image prediction is provided via a voxel-vote method using overlapping windows. Our method produces parcellated images that are highly consistent with standard atlas-based methods in under 3 min on a modern GPU, and the single network architecture allows for rapid retraining (<36 hr) as needed.

Construction of immune-related molecular diagnostic and predictive models of hepatocellular carcinoma based on machine learning

Background

To exploit hepatocellular carcinoma (HCC) diagnostic substances, we identify potential predictive markers based on machine learning and to explore the significance of immune cell infiltration in this pathology.

Method

Three HCC gene expression datasets were used for weighted gene co-expression network analysis (WGCNA) and differential expression analysis. Least Absolute Shrinkage and Selection Operator (LASSO) and Random Forest were applied to identify candidate biomarkers. The diagnostic value of HCC diagnostic gene biomarkers was further assessed by the area under the ROC curve observed in the validation dataset. CIBERSORT was used to analyze 22 immune cell fractions from HCC patients and to analyze their correlation with diagnostic markers. In addition, the prognostic value of the markers and the sensitivity of the drugs were analyzed.

Result

WGCNA and differential expression analysis were used to screen 396 distinct gene signatures in HCC tissues. They were mostly engaged in cytoplasmic fusion and the cell division cycle, according to gene enrichment analyses. Five genes were shown to have a high diagnostic value for use as diagnostic biomarkers for HCC, including EFHD1 (AUC = 0.77), KIF4A (AUC = 0.97), UBE2C (AUC = 0.96), SMYD3 (AUC = 0.91), and MCM7 (AUC = 0.93). T cells, NK cells, macrophages, and dendritic cells were found to be related to diagnostic markers in HCC tissues by immune cell infiltration analysis, indicating that these cells are intimately linked to the onset and spread of HCC. Concurrently, these five genes and their constructed models have considerable prognostic value.

Conclusion

These five genes (EFHD1, KIF4A, UBE2C, SMYD3, and MCM7) may serve as new candidate molecular markers for HCC, providing new insights for future diagnosis, prognosis, and molecular therapy of HCC.

S100A9 Induces Macrophage M2 Polarization and Immunomodulatory Role in the Lesion Site After Spinal Cord Injury in Rats

Immune response is pivotal in the secondary injury of spinal cord injury (SCI). Polarization of macrophages (MΦ) influences the immune response in the secondary injury, which is regulated by several immune-related proteins. M2Φ plays the immunomodulatory role in the central nervous system. This study used bioinformatic analysis and machine algorithms to screen hub immune-related proteins after SCI and experimentally investigate the role of the target protein in the M2Φ polarization and immunomodulation in rats and in vitro after SCI. We downloaded GSE151371 and GSE45006, hub immune-related genes were screened using machine learning algorithms, and the expression of S100A9 was verified by datasets. Allen's weight-drop injury SCI model in Sprague-Dawley rat and bone marrow-derived rat MΦ with myelin debris model were used to study the effects of S100A9 on M2Φ polarization and immunomodulation at the lesion site and in vitro. Bioinformatic analysis showed that S100A9 acts as a hub immune-related gene in the SCI patients and rats. S100A9 increased at the lesion site in SCI rats, and its inhibition reduced CD206 and ARG-1 expression. Exogenous S100A9 promoted CD206 and ARG-1 expression in MΦ. S100A9 also increased the expression of PD-L1 and decreased MHC II at the lesion site in SCI rats and MΦ with myelin debris, and enhanced mitochondrial activity in rat MΦ with myelin debris. In conclusion, S100A9 is an indispensable factor in the immune process in secondary injury following SCI.

Using meta-learning to recommend an appropriate time-series forecasting model

BACKGROUND

There are various forecasting algorithms available for univariate time series, ranging from simple to sophisticated and computational. In practice, selecting the most appropriate algorithm can be difficult, because there are too many algorithms. Although expert knowledge is required to make an informed decision, sometimes it is not feasible due to the lack of such resources as time, money, and manpower.

METHODS

In this study, we used coronavirus disease 2019 (COVID-19) data, including the absolute numbers of confirmed, death and recovered cases per day in 187 countries from February 20, 2020, to May 25, 2021. Two popular forecasting models, including Auto-Regressive Integrated Moving Average (ARIMA) and exponential smoothing state-space model with Trigonometric seasonality, Box-Cox transformation, ARMA errors, Trend, and Seasonal components (TBATS) were used to forecast the data. Moreover, the data were evaluated by the root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and symmetric mean absolute percentage error (SMAPE) criteria to label time series. The various characteristics of each time series based on the univariate time series structure were extracted as meta-features. After that, three machine-learning classification algorithms, including support vector machine (SVM), decision tree (DT), random forest (RF), and artificial neural network (ANN) were used as meta-learners to recommend an appropriate forecasting model.

RESULTS

The finding of the study showed that the DT model had a better performance in the classification of time series. The accuracy of DT in the training and testing phases was 87.50% and 82.50%, respectively. The sensitivity of the DT algorithm in the training phase was 86.58% and its specificity was 88.46%. Moreover, the sensitivity and specificity of the DT algorithm in the testing phase were 73.33% and 88%, respectively.

CONCLUSION

In general, the meta-learning approach was able to predict the appropriate forecasting model (ARIMA and TBATS) based on some time series features. Considering some characteristics of the desired COVID-19 time series, the ARIMA or TBATS forecasting model might be recommended to forecast the death, confirmed, and recovered trend cases of COVID-19 by the DT model.

Rationale and design of the CONFIRM2 (Quantitative COroNary CT Angiography Evaluation For Evaluation of Clinical Outcomes: An InteRnational, Multicenter Registry) study

BACKGROUND

In the last 15 years, large registries and several randomized clinical trials have demonstrated the diagnostic and prognostic value of coronary computed tomography angiography (CCTA). Advances in CT scanner technology and developments of analytic tools now enable accurate quantification of coronary artery disease (CAD), including total coronary plaque volume and low attenuation plaque volume. The primary aim of CONFIRM2, (Quantitative COroNary CT Angiography Evaluation For Evaluation of Clinical Outcomes: An InteRnational, Multicenter Registry) is to perform comprehensive quantification of CCTA findings, including coronary, non-coronary cardiac, non-cardiac vascular, non-cardiac findings, and relate them to clinical variables and cardiovascular clinical outcomes.

DESIGN

CONFIRM2 is a multicenter, international observational cohort study designed to evaluate multidimensional associations between quantitative phenotype of cardiovascular disease and future adverse clinical outcomes in subjects undergoing clinically indicated CCTA. The targeted population is heterogenous and includes patients undergoing CCTA for atherosclerotic evaluation, valvular heart disease, congenital heart disease or pre-procedural evaluation. Automated software will be utilized for quantification of coronary plaque, stenosis, vascular morphology and cardiac structures for rapid and reproducible tissue characterization. Up to 30,000 patients will be included from up to 50 international multi-continental clinical CCTA sites and followed for 3-4 years.

SUMMARY

CONFIRM2 is one of the largest CCTA studies to establish the clinical value of a multiparametric approach to quantify the phenotype of cardiovascular disease by CCTA using automated imaging solutions.

Airway Cells 3D Reconstruction via Manual and Machine-Learning Aided Segmentation of Volume EM Datasets

Volume electron microscopy (vEM) is a high-resolution imaging technique capable of revealing the 3D structure of cells, tissues, and model organisms. This imaging modality is gaining prominence due to its ability to provide a comprehensive view of cells at the nanometer scale. The visualization and quantitative analysis of individual subcellular structures however requires segmentation of each 2D electron micrograph slice of the 3D vEM dataset; this process is extremely laborious de facto limiting its applications and throughput. To address these limitations, deep learning approaches have been recently developed including Empanada-Napari plugin, an open-source tool for automated segmentation based on a Panoptic-DeepLab (PDL) architecture. In this chapter, we provide a step-by-step protocol describing the process of manual segmentation using 3dMOD within the IMOD package and the process of automated segmentation using Empanada-Napari plugins for the 3D reconstruction of airway cellular structures.

Ultra-low esfenvalerate exposure may disrupt interspecific competition

Field and mesocosm studies repeatedly show that higher tier processes reduce the predictive accuracy of toxicity evaluation and thus their value for pesticide risk assessment. Therefore, understanding the influence of ecological complexity on toxicant effects is crucial to improve realism of aquatic risk assessment. Here we investigate the influence of repeated exposure to ecologically realistic concentrations of esfenvalerate on the two similarly sensitive species Daphnia magna and Culex pipiens in a food limited and highly competitive environment. We show that significant perturbations in population development are only present at 100 ng/L (close to the EC50). In contrast, interspecific competition between species is already reduced at 0.1 ng/L (≤ 3 orders of magnitude below the acute lethal EC50). We conclude that extremely low, environmentally relevant concentrations can disrupt species interactions. This toxicant mediated alteration of competitive balances in ecological communities may be the underlying mechanism for shifts in species distribution at ultra-low pesticide concentrations. A realistic risk assessment should therefore consider these processes in order to predict potential pesticide effects on the structure of communities.

Computational approaches for identifying disease-causing mutations in proteins

Advancements in genome sequencing have expanded the scope of investigating mutations in proteins across different diseases. Amino acid mutations in a protein alter its structure, stability and function and some of them lead to diseases. Identification of disease-causing mutations is a challenging task and it will be helpful for designing therapeutic strategies. Hence, mutation data available in the literature have been curated and stored in several databases, which have been effectively utilized for developing computational methods to identify deleterious mutations (drivers), using sequence and structure-based properties of proteins. In this chapter, we describe the contents of specific databases that have information on disease-causing and neutral mutations followed by sequence and structure-based properties. Further, characteristic features of disease-causing mutations will be discussed along with computational methods for identifying cancer hotspot residues and disease-causing mutations in proteins.

Development and validation of a machine learning-based fall-related injury risk prediction model using nationwide claims database in Korean community-dwelling older population

BACKGROUND

Falls impact over 25% of older adults annually, making fall prevention a critical public health focus. We aimed to develop and validate a machine learning-based prediction model for serious fall-related injuries (FRIs) among community-dwelling older adults, incorporating various medication factors.

METHODS

Utilizing annual national patient sample data, we segmented outpatient older adults without FRIs in the preceding three months into development and validation cohorts based on data from 2018 and 2019, respectively. The outcome of interest was serious FRIs, which we defined operationally as incidents necessitating an emergency department visit or hospital admission, identified by the diagnostic codes of injuries that are likely associated with falls. We developed four machine-learning models (light gradient boosting machine, Catboost, eXtreme Gradient Boosting, and Random forest), along with a logistic regression model as a reference.

RESULTS

In both cohorts, FRIs leading to hospitalization/emergency department visits occurred in approximately 2% of patients. After selecting features from initial set of 187, we retained 26, with 15 of them being medication-related. Catboost emerged as the top model, with area under the receiver operating characteristic of 0.700, along with sensitivity and specificity rates around 65%. The high-risk group showed more than threefold greater risk of FRIs than the low-risk group, and model interpretations aligned with clinical intuition.

CONCLUSION

We developed and validated an explainable machine-learning model for predicting serious FRIs in community-dwelling older adults. With prospective validation, this model could facilitate targeted fall prevention strategies in primary care or community-pharmacy settings.

Facial profile evaluation and prediction of skeletal class II patients during camouflage extraction treatment: a pilot study

BACKGROUND

The evaluation of the facial profile of skeletal Class II patients with camouflage treatment is of great importance for patients and orthodontists. The aim of this study is to explore the key factors in evaluating the facial profile esthetics and to predict the posttreatment facial profile esthetics of skeletal Class II extraction patients.

METHODS

124 skeletal Class II extraction patients were included. The pretreatment and posttreatment cephalograms were analyzed by a trained expert orthodontist. The facial profile esthetics of pretreatment and posttreatment lateral photographs were evaluated by 10 expert orthodontists using the visual analog scale (VAS). The correlation between subjective facial profile esthetics and objective cephalometric measurements was assessed. Three machine-learning methods were used to predict posttreatment facial profile esthetics.

RESULTS

The distances from lower and upper lip to the E plane and U1-APo showed the stronger correlation with profile esthetics. The changes in lower lip to the E plane and U1-APo during extraction exhibited the stronger correlation with changes in VAS score (r = - 0.551 and r = - 0.469). The random forest prediction model had the lowest mean absolute error and root mean square error, demonstrating a better prediction accuracy and fitting effect. In this model, pretreatment upper lip to E plane, pretreatment Pog-NB and the change of U1-GAll were the most important variables in predicting the posttreatment score of facial profile esthetics.

CONCLUSIONS

The maxillary incisor protrusion and lower lip protrusion are key objective indicators for evaluating and predicting facial profile esthetics of skeletal Class II extraction patients. An artificial intelligence prediction model could be a new method for predicting the posttreatment esthetics of facial profiles.

Measuring dairy cow welfare with real-time sensor-based data and farm records: a concept study

Welfare assessment of dairy cows by in-person farm visits provides only a snapshot of welfare and is time-consuming and costly. Possible solutions to reduce the need for in-person assessments would be to exploit sensor data and other routinely collected on-farm records. The aim of this study was to develop an algorithm to classify dairy cow welfare based on sensors (accelerometer and/or milk meter) and farm records (e.g. days in milk, lactation number). In total, 318 cows from six commercial farms located in Finland, Italy and Spain (two farms each) were enrolled for a pilot study lasting 135 days. During this time, cows were routinely scored using 14 animal-based measures of good feeding, health and housing based on the Welfare Quality® (WQ®) protocol. WQ® measures were evaluated daily or approximately every 45 days, using disease treatments from farm records and on-farm visits, respectively. WQ® measures were supplemented with daily temperature-humidity index to account for heat stress. The severity and duration of each welfare measure were evaluated, and the final welfare index was obtained by summing up the values for each cow on each pilot study day, and stratifying the result into three classes: good, moderate and poor welfare. For model building, a machine-learning (ML) algorithm based on gradient-boosted trees (XGBoost) was applied. Two model versions were tested: (1) a global model tested on unseen herd, and (2) a herd-specific model tested on unseen part of the data from the same herd. The version (1) served as an example on the model performance on a herd not previsited by the evaluator, while version (2) resembled a custom-made solution requiring in-person welfare evaluation for model training. Our results indicated that the global model had a low performance with average sensitivity and specificity of 0.44 and 0.68, respectively. For the herd-specific version, the model performance was higher reaching an average of 0.64 sensitivity and 0.80 specificity. The highest classification performance was obtained for cows in poor welfare, followed by cows in good and moderate welfare (balanced accuracy of 0.77, 0.71 and 0.68, respectively). Since the global model had low classification accuracy, the use of the developed model as a stand-alone system based solely on sensor data is infeasible, and a combination of in-person and sensor-based welfare evaluation would be preferable for a reliable welfare assessment. ML-based solutions, even with fair discriminative abilities, have the potential to enhance dairy welfare monitoring.

Multiphase diets may improve feed efficiency in fattening crossbreed Holstein bulls: a retrospective simulation of the economic and environmental impact

Beef industry needs alternative feeding strategies to enhance both economic and environmental sustainability. Among these strategies, adjusting the diet dynamically according to the change of nutritional requirements (multiphase diet) has demonstrated its economic and environmental benefits in pig production systems. Therefore, this retrospective study aims to assess, through simulation, the theoretical economic and environmental benefits of introducing a multiphase diet for crossbreed bulls feeding (one or more diet changes). For this, individual data of BW, BW gain, and daily intake were recorded from 342 bulls during the last fattening period (112 days). These data were used to estimate individual trajectory of energy and protein requirements, which were subsequently divided by individual intake to calculate the required dietary energy and protein concentrations. The area between two functions (i.e., ƒ1: constant protein concentration in the original diet during fattening and ƒ2: estimated protein concentration requirements) was minimised to identify the optimal moments to adjust the dietary concentration of energy and protein. The results indicated that both energy and protein intake exceeded requirements on average (+16% and +28% respectively, P < 0.001), justifying the adoption of a multiphase diet. Modelling the individual trajectories of required metabolisable protein (MP, g/kg DM) during the fattening period resulted in exponential decay model in relation to BW [32120 × exp(-0.026 × BW) + 59.9], while the dietary net energy concentration followed a slightly quadratic model [2.26-0.0026 × BW + 0.000003 × BW2]. Minimisation of the area between curves showed two optimal moments to adjust the diet: at 312 kg and 385 kg of BW, indicating three diet phases: (a) <312 kg, (b) 312-385 kg, and (c) 385-600 kg. For the second and third phases, the dietary energy and protein concentration should be 70 g MP/kg DM and 1.70 Mcal/kg DM and 61 g MP/kg DM and 1.65 Mcal/kg DM, respectively. These diet adjustments might improve economic profitability by 29 €/animal, reduce estimated nitrogen excretions by 16% (P < 0.001), and maintain similar weight gain (P > 0.16) compared to the commercial diet. However, the decrease in dietary energy concentration led to increased fibre concentration, which in turn increased the estimated CH4 emissions of animals with the multiphase diet (+44%, P < 0.001). Hence, multiphase diet could theoretically reduce feeding cost and nitrogen excretion from fattening cattle. Further in vivo studies should confirm these results and find optimal nutritional strategies to improve economic profitability and environmental impact.

Opening the Black Box of Family-Based Treatments: An Artificial Intelligence Framework to Examine Therapeutic Alliance and Therapist Empathy

The evidence-based treatment (EBT) movement has primarily focused on core intervention content or treatment fidelity and has largely ignored practitioner skills to manage interpersonal process issues that emerge during treatment, especially with difficult-to-treat adolescents (delinquent, substance-using, medical non-adherence) and those of color. A chief complaint of "real world" practitioners about manualized treatments is the lack of correspondence between following a manual and managing microsocial interpersonal processes (e.g. negative affect) that arise in treating "real world clients." Although family-based EBTs share core similarities (e.g. focus on family interactions, emphasis on practitioner engagement, family involvement), most of these treatments do not have an evidence base regarding common implementation and treatment process problems that practitioners experience in delivering particular models, especially in mid-treatment when demands on families to change their behavior is greatest in treatment - a lack that characterizes the field as a whole. Failure to effectively address common interpersonal processes with difficult-to-treat families likely undermines treatment fidelity and sustained use of EBTs, treatment outcome, and contributes to treatment dropout and treatment nonadherence. Recent advancements in wearables, sensing technologies, multivariate time-series analyses, and machine learning allow scientists to make significant advancements in the study of psychotherapy processes by looking "under the skin" of the provider-client interpersonal interactions that define therapeutic alliance, empathy, and empathic accuracy, along with the predictive validity of these therapy processes (therapeutic alliance, therapist empathy) to treatment outcome. Moreover, assessment of these processes can be extended to develop procedures for training providers to manage difficult interpersonal processes while maintaining a physiological profile that is consistent with astute skills in psychotherapeutic processes. This paper argues for opening the "black box" of therapy to advance the science of evidence-based psychotherapy by examining the clinical interior of evidence-based treatments to develop the next generation of audit- and feedback- (i.e., systemic review of professional performance) supervision systems.

Interactive polar diagrams for model comparison

OBJECTIVE

Evaluating the performance of multiple complex models, such as those found in biology, medicine, climatology, and machine learning, using conventional approaches is often challenging when using various evaluation metrics simultaneously. The traditional approach, which relies on presenting multi-model evaluation scores in the table, presents an obstacle when determining the similarities between the models and the order of performance.

METHODS

By combining statistics, information theory, and data visualization, juxtaposed Taylor and Mutual Information Diagrams permit users to track and summarize the performance of one model or a collection of different models. To uncover linear and nonlinear relationships between models, users may visualize one or both charts.

RESULTS

Our library presents the first publicly available implementation of the Mutual Information Diagram and its new interactive capabilities, as well as the first publicly available implementation of an interactive Taylor Diagram. Extensions have been implemented so that both diagrams can display temporality, multimodality, and multivariate data sets, and feature one scalar model property such as uncertainty. Our library, named polar-diagrams, supports both continuous and categorical attributes.

CONCLUSION

The library can be used to quickly and easily assess the performances of complex models, such as those found in machine learning, climate, or biomedical domains.

Covariate dependent Markov chains constructed with gradient boost modeling can effectively generate long-term predictions of obesity trends

IMPORTANCE

The prevalence of obesity among United States adults has increased from 30.5% in 1999 to 41.9% in 2020. However, despite the recognition of long-term weight gain as an important public health issue, there is a paucity of studies studying the long-term weight gain and building models for long-term projection.

METHODS

A retrospective, cross-sectional cohort study using the publicly available National Health and Nutrition Examination Survey (NHANES 2017-2020) was conducted in patients who completed the weight questionnaire and had accurate data for both weight at time of survey and weight ten years ago. Multistate gradient boost modeling classifiers were used to generate covariate dependent transition matrices and Markov chains were utilized for multistate modeling.

RESULTS

Of the 6146 patients that met the inclusion criteria, 3024 (49%) of patients were male and 3122 (51%) of patients were female. There were 2252 (37%) White patients, 1257 (20%) Hispanic patients, 1636 (37%) Black patients, and 739 (12%) Asian patients. The average BMI was 30.16 (SD = 7.15), the average weight was 83.67 kilos (SD = 22.04), and the average weight change was a 3.27 kg (SD = 14.97) increase in body weight (Fig. 1). A total of 2411 (39%) patients lost weight, and 3735 (61%) patients gained weight (Table 1). We observed that 87 (1%) of patients were underweight (BMI < 18.5), 2058 (33%) were normal weight (18.5 ≤ BMI < 25), 1376 (22%) were overweight (25 ≤ BMI < 30) and 2625 (43%) were obese (BMI > 30). From analysis of the transitions between normal/underweight, overweight, and obese, we observed that after 10 years, of the patients who were underweight, 65% stayed underweight, 32% became normal weight, 2% became overweight, and 2% became obese. After 10 years, of the patients who were normal weight, 3% became underweight, 78% stayed normal weight, 17% became overweight, and 2% became obese. Of the patients who were overweight, 71% stayed overweight, 0% became underweight, 14% became normal weight, and 15% became obese. Of the patients who were obese, 84% stayed obese, 0% became underweight, 1% became normal weight, and 14% became overweight.

CONCLUSIONS

United States adults are at risk of transitioning from normal weight to becoming overweight or obese. Covariate dependent Markov chains constructed with gradient boost modeling can effectively generate long-term predictions.

Should All Pancreatic Cystic Lesions with Worrisome or High-Risk Features Be Resected? A Clinical and Radiological Machine Learning Model May Help to Answer

RATIONALE AND OBJECTIVES

According to current guidelines, pancreatic cystic lesions (PCLs) with worrisome or high-risk features may have overtreatment. The purpose of this study was to build a clinical and radiological based machine-learning (ML) model to identify malignant PCLs for surgery among preoperative PCLs with worrisome or high-risk features.

MATERIALS AND METHODS

Clinical and radiological details of 317 pathologically confirmed PCLs with worrisome or high-risk features were retrospectively analyzed and applied to ML models including Support Vector Machine, Logistic Regression (LR), Decision Tree, Bernoulli NB, Gaussian NB, K Nearest Neighbors and Linear Discriminant Analysis. The diagnostic ability for malignancy of the optimal model with the highest diagnostic AUC in the cross-validation procedure was further evaluated in internal (n = 77) and external (n = 50) testing cohorts, and was compared to two published guidelines in internal mucinous cyst cohort.

RESULTS

Ten clinical and radiological feature-based LR model was the optimal model with the highest AUC (0.951) in the cross-validation procedure. In the internal testing cohort, LR model reached an AUC, accuracy, sensitivity, and specificity of 0.927, 0.909, 0.914, and 0.905; in the external testing cohort, LR model reached 0.948, 0.900, 0.963, and 0.826. When compared to the European guidelines and the ACG guidelines, LR model demonstrated significantly better accuracy and specificity in identifying malignancy, while maintaining the same high sensitivity.

CONCLUSION

Clinical- and radiological-based LR model can accurately identify malignant PCLs in patients with worrisome or high-risk features, possessing diagnostic performance better than the European guidelines as well as ACG guidelines.

HIV-1 subtypes maintain distinctive physicochemical signatures in Nef domains associated with immunoregulation

BACKGROUND

HIV subtype is associated with varied rates of disease progression. The HIV accessory protein, Nef, continues to be present during antiretroviral therapy (ART) where it has numerous immunoregulatory effects. In this study, we analyzed Nef sequences from HIV subtypes A1, B, C, and D using a machine learning approach that integrates functional amino acid information to identify if unique physicochemical features are associated with Nef functional/structural domains in a subtype-specific manner.

METHODS

2253 sequences representing subtypes A1, B, C, and D were aligned and domains with known functional properties were scored based on amino acid physicochemical properties. Following feature generation, we used statistical pruning and evolved neural networks (ENNs) to determine if we could successfully classify subtypes. Next, we used ENNs to identify the top five key Nef physicochemical features applied to specific immunoregulatory domains that differentiated subtypes. A signature pattern analysis was performed to the assess amino acid diversity in sub-domains that differentiated each subtype.

RESULTS

In validation studies, ENNs successfully differentiated each subtype at A1 (87.2%), subtype B (89.5%), subtype C (91.7%), and subtype D (85.1%). Our feature-based domain scoring, followed by t-tests, and a similar ENN identified subtype-specific domain-associated features. Subtype A1 was associated with alterations in Nef CD4 binding domain; subtype B was associated with alterations with the AP-2 Binding domain; subtype C was associated with alterations in a structural Alpha Helix domain; and, subtype D was associated with alterations in a Beta-Sheet domain.

CONCLUSIONS

Recent studies have focused on HIV Nef as a driver of immunoregulatory disease in those HIV infected and on ART. Nef acts through a complex mixture of interactions that are directly linked to the key features of the subtype-specific domains we identified with the ENN. The study supports the hypothesis that varied Nef subtypes contribute to subtype-specific disease progression.

Machine-learning and combined analysis of single-cell and bulk-RNA sequencing identified a DC gene signature to predict prognosis and immunotherapy response for patients with lung adenocarcinoma

BACKGROUND

Innate immune effectors, dendritic cells (DCs), influence cancer prognosis and immunotherapy significantly. As such, dendritic cells are important in killing tumors and influencing tumor microenvironment, whereas their roles in lung adenocarcinoma (LUAD) are largely unknown.

METHODS

In this study, 1658 LUAD patients from different cohorts were included. In addition, 724 cancer patients who received immunotherapy were also included. To identify DC marker genes in LUAD, we used single-cell RNAsequencing data for analysis and determined 83 genes as DC marker genes. Following that, integrative machine learning procedure was developed to construct a signature for DC marker genes.

RESULTS

Using TCGA bulk-RNA sequencing data as the training set, we developed a signature consisting of seven genes and classified patients by their risk status. Another six independent cohorts demonstrated the signature' s prognostic power, and multivariate analysis demonstrated it was an independent prognostic factor. LUAD patients in the high-risk group displayed more advanced features, discriminatory immune-cell infiltrations and immunosuppressive states. Cell-cell communication analysis indicates that tumor cells with lower risk scores communicate more actively with the tumor microenvironment. Eight independent immunotherapy cohorts revealed that patients with low-risk had better immunotherapy responses. Drug sensitivity analysis indicated that targeted therapy agents exhibited greater sensitivity to low-risk patients, while chemotherapy agents displayed greater sensitivity to high-risk patients. In vitro experiments confirmed that CTSH is a novel protective factor for LUAD.

CONCLUSIONS

An unique signature based on DC marker genes that is highly predictive of LUAD patients' prognosis and response to immunotherapy. CTSH is a new biomarker for LUAD.

Translation of Machine Learning-Based Prediction Algorithms to Personalised Empiric Antibiotic Selection: A Population-Based Cohort Study

BACKGROUND

Prediction of antibiotic non-susceptibility based on patient characteristics and clinical status may support selection of empiric antibiotics for suspected hospital-acquired urinary tract infections (HA-UTIs).

METHODS

Prediction models were developed to predict non-susceptible results of eight antibiotic susceptibility tests ordered for suspected HA-UTI. Eligible patients were those with urine culture and susceptibility test results after 48 hours of admission between 2010-2021. Patient demographics, diagnosis, prescriptions, exposure to multidrug-resistant organisms, transfer history, and a daily calculated antibiogram were used as predictors. Lasso logistic regression (LLR), extreme gradient boosting (XGB), random forest, and stacked ensemble methods were used for development. Parsimonious models were also developed for clinical utility. Discrimination was assessed using the area under the receiver operating characteristic curve (AUROC).

RESULTS

In 10 474 suspected HA-UTI cases, the mean age was 62.1 ± 16.2 years and 48.1% were male. Non-susceptibility prediction for ampicillin/sulbactam, cefepime, ciprofloxacin, imipenem, piperacillin/tazobactam, and trimethoprim/sulfamethoxazole performed best using the stacked ensemble (AUROC 76.9, 76.1, 77.0, 80.6, 76.1, and 76.5, respectively). The model for ampicillin performed best with LLR (AUROC 73.4). Extreme gradient boosting only performed best for gentamicin (AUROC 66.9). In the parsimonious models, the LLR yielded the highest AUROC for ampicillin, ampicillin/sulbactam, cefepime, gentamicin, and trimethoprim/sulfamethoxazole (AUROC 70.6, 71.8, 73.0, 65.9, and 73.0, respectively). The model for ciprofloxacin performed best with XGB (AUROC 70.3), while the model for imipenem performed best in the stacked ensemble (AUROC 71.3). A personalised application using the parsimonious models was publicly released.

CONCLUSIONS

Prediction models for antibiotic non-susceptibility were developed to support empiric antibiotic selection for HA-UTI.

A tumor-infiltrating immune cells-related pseudogenes signature based on machine-learning predicts outcomes and immunotherapy responses in ovarian cancer

Previous researches have provided evidence for the significant involvement of pseudogenes in immune-related functions across different types of cancer. However, the mechanisms by which pseudogenes regulate immunity in ovarian cancer (OV) and their potential impact on clinical outcomes remain unclear. To address this gap in knowledge, our study utilized a novel computational framework to analyze a total of 491 samples from three public datasets. We employed a combination of 10 machine-learning algorithms to construct a signature known as the tumor-infiltrating immune cells-related pseudogenes signature (TIICPS). The TIICPS, consisting of 12 pseudogenes, demonstrated independent prognostic value for overall survival, surpassing conventional clinical traits, 62 published signatures, and TP53 and BRCA mutation status in three cohorts. Patients with low TIICPS exhibited heightened immune-related pathways, intricate genomic alterations, substantial immune infiltration, and greater potential for immunotherapy efficacy. Consequently, TIICPS holds promise as a predictive tool for prognosis and immunotherapy response in ovarian cancer.

Toward practical machine-learning-based diagnosis for drug-naïve women with major depressive disorder using EEG channel reduction approach

BACKGROUND

A machine-learning-based computer-aided diagnosis (CAD) system can complement the traditional diagnostic error for major depressive disorder (MDD) using trait-like neurophysiological biomarkers. Previous studies have shown that the CAD system has the potential to differentiate between female MDD patients and healthy controls. The aim of this study was to develop a practically useful resting-state electroencephalography (EEG)-based CAD system to assist in the diagnosis of drug-naïve female MDD patients by considering both the drug and gender effects. In addition, the feasibility of the practical use of the resting-state EEG-based CAD system was evaluated using a channel reduction approach.

METHODS

Eyes-closed, resting-state EEG data were recorded from 49 drug-naïve female MDD patients and 49 sex-matched healthy controls. Six different EEG feature sets were extracted: power spectrum densities (PSDs), phase-locking values (PLVs), and network indices for both sensor- and source-level, and four different EEG channel montages (62, 30, 19, and 10-channels) were designed to investigate the channel reduction effects in terms of classification performance.

RESULTS

The classification performances for each feature set were evaluated using a support vector machine with leave-one-out cross-validation. The optimum classification performance was achieved when using sensor-level PLVs (accuracy: 83.67 % and area under curve: 0.92). Moreover, the classification performance was maintained until the number of EEG channels was reduced to 19 (over 80 % accuracy).

CONCLUSION

We demonstrated the promising potential of sensor-level PLVs as diagnostic features when developing a resting-state EEG-based CAD system for the diagnosis of drug-naïve female MDD patients and verified the feasibility of the practical use of the developed resting-state EEG-based CAD system using the channel reduction approach.

Assessment and risk prediction of frailty using texture-based muscle ultrasound image analysis and machine learning techniques

The purpose of this study was to evaluate texture-based muscle ultrasound image analysis for the assessment and risk prediction of frailty phenotype. This retrospective study of prospectively acquired data included 101 participants who underwent ultrasound scanning of the anterior thigh. Participants were subdivided according to frailty phenotype and were followed up for two years. Primary and secondary outcome measures were death and comorbidity, respectively. Forty-three texture features were computed from the rectus femoris and the vastus intermedius muscles using statistical methods. Model performance was evaluated by computing the area under the receiver operating characteristic curve (AUC) while outcome prediction was evaluated using regression analysis. Models developed achieved a moderate to good AUC (0.67 ≤ AUC ≤ 0.79) for categorizing frailty. The stepwise multiple logistic regression analysis demonstrated that they correctly classified 70-87% of the cases. The models were associated with increased comorbidity (0.01 ≤ p ≤ 0.18) and were predictive of death for pre-frail and frail participants (0.001 ≤ p ≤ 0.016). In conclusion, texture analysis can be useful to identify frailty and assess risk prediction (i.e. mortality) using texture features extracted from muscle ultrasound images in combination with a machine learning approach.

Improved estimation of the risk of manic relapse by combining clinical and brain scan data

INTRODUCTION

Estimating the risk of manic relapse could help the psychiatrist individually adjust the treatment to the risk. Some authors have attempted to estimate this risk from baseline clinical data. Still, no studies have assessed whether the estimation could improve by adding structural magnetic resonance imaging (MRI) data. We aimed to evaluate it.

MATERIAL AND METHODS

We followed a cohort of 78 patients with a manic episode without mixed symptoms (bipolar type I or schizoaffective disorder) at 2-4-6-9-12-15-18 months and up to 10 years. Within a cross-validation scheme, we created and evaluated a Cox lasso model to estimate the risk of manic relapse using both clinical and MRI data.

RESULTS

The model successfully estimated the risk of manic relapse (Cox regression of the time to relapse as a function of the estimated risk: hazard ratio (HR)=2.35, p=0.027; area under the curve (AUC)=0.65, expected calibration error (ECE)<0.2). The most relevant variables included in the model were the diagnosis of schizoaffective disorder, poor impulse control, unusual thought content, and cerebellum volume decrease. The estimations were poorer when we used clinical or MRI data separately.

CONCLUSION

Combining clinical and MRI data may improve the risk of manic relapse estimation after a manic episode. We provide a website that estimates the risk according to the model to facilitate replication by independent groups before translation to clinical settings.

Machine-learning-based prediction of fractional flow reserve after percutaneous coronary intervention

BACKGROUND AND AIMS

Post-percutaneous coronary intervention (PCI) fractional flow reserve (FFR) reflects residual atherosclerotic burden and is associated with future events. How much post-PCI FFR can be predicted based on baseline basic information and the clinical relevance have not been investigated.

METHODS

We compiled a multicenter registry of patients undergoing pre- and post-PCI FFR. Machine-learning (ML) algorithms were designed to predict post-PCI FFR levels from baseline demographics, quantitative coronary angiography, and pre-PCI FFR. FFR deviation was defined as actual minus ML-predicted post-PCI FFR levels, and its association with incident target vessel failure (TVF) was evaluated.

RESULTS

Median (IQR) pre- and post-PCI FFR values were 0.71 (0.61, 0.77) and 0.88 (0.84, 0.93), respectively. The Spearman correlation coefficient of the actual and predicted post-PCI FFR was 0.54 (95% CI: 0.52, 0.57). FFR deviation was non-linearly associated with incident TVF (HR [95% CI] with Q3 as reference: 1.65 [1.14, 2.39] in Q1, 1.42 [0.98, 2.08] in Q2, 0.81 [0.53, 1.26] in Q4, and 1.04 [0.69, 1.56] in Q5). A model with polynomial function of continuous FFR deviation indicated increasing TVF risk for FFR deviation ≤0 but plateau risk with FFR deviation >0.

CONCLUSIONS

An ML-based algorithm using baseline data moderately predicted post-PCI FFR. The deviation of post-PCI FFR from the predicted value was associated with higher vessel-oriented event.

Location-specific ASPECTS does not improve Outcome Prediction in Large Vessel Occlusion compared to Cumulative ASPECTS

PURPOSE

Individual regions of the Alberta Stroke Programme Early CT Score (ASPECTS) may contribute differently to the clinical symptoms in large vessel occlusion (LVO). Here, we investigated whether the predictive performance on clinical outcome can be increased by considering specific ASPECTS subregions.

METHODS

A consecutive series of patients with LVO affecting the middle cerebral artery territory and subsequent endovascular treatment (EVT) between January 2015 and July 2020 was analyzed, including affected ASPECTS regions. A multivariate logistic regression was performed to assess the individual impact of ASPECTS regions on good clinical outcome (defined as modified Rankin scale after 90 days of 0-2). Machine-learning-driven logistic regression models were trained (training = 70%, testing = 30%) to predict good clinical outcome using i) cumulative ASPECTS and ii) location-specific ASPECTS, and their performance compared using deLong's test. Furthermore, additional analyses using binarized as well as linear clinical outcomes using regression and machine-learning techniques were applied to thoroughly assess the potential predictive properties of individual ASPECTS regions and their combinations.

RESULTS

Of 1109 patients (77.3 years ± 11.6, 43.8% male), 419 achieved a good clinical outcome and a median NIHSS after 24 h of 12 (interquartile range, IQR 4-21). Individual ASPECTS regions showed different impact on good clinical outcome in the multivariate logistic regression, with strongest effects for insula (odds ratio, OR 0.56, 95% confidence interval, CI 0.42-0.75) and M5 (OR 0.53, 95% CI 0.29-0.97) regions. Accuracy (ACC) in predicting good clinical outcome of the test set did not differ between when considering i) cumulative ASPECTS and ii) location-specific ASPECTS (ACC = 0.619, 95% CI 0.58-0.64 vs. ACC = 0.629, 95% CI 0.60-0.65; p = 0.933).

CONCLUSION

Cumulative ASPECTS assessment in LVO remains a stable and reliable predictor for clinical outcome and is not inferior to a weighted (location-specific) ASPECTS assessment.

LiRA-CD: An open-source dataset for road condition modelling and research

This data article presents the details of the Live Road Assessment Custom Dataset (LiRA-CD), an open-source dataset for road condition modelling and research. The dataset captures GPS trajectories of a fleet of electric vehicles and their time-series data from 50 different sensors collected on 230 km of highway and urban roads in Copenhagen, Denmark. Additionally, road condition measurements were collected by standard survey vehicles, which serve as high-quality reference data. The in-vehicle measurements were collected onboard with an Internet-of-Things (IoT) device, then periodically transmitted before being saved in a database. Researchers can use the dataset for prediction modelling related to standard road condition parameters such as surface friction and texture, road roughness, road damages, and energy consumption. Furthermore, researchers and pavement engineers can use the dataset as a template for future studies and projects, benchmarking the performance of different algorithms and solving problems of the same type. LiRA-CD is freely available and can be accessed at https://doi.org/10.11583/DTU.c.6659909.

Identification of the Key Immune Cells and Genes for the Diagnostics and Therapeutics of Meningioma

BACKGROUND

Dysregulation of immune infiltration critically contributes to the tumorigenesis and progression of meningiomas. However, the landscape of immune microenvironment and key genes correlated with immune cell infiltration remains unclear.

METHODS

Four Gene Expression Omnibus data sets were included. CIBERSORT algorithm was utilized to analyze the immune cell infiltration in samples. Wilcoxon test, Random Forest algorithm, and Least Absolute Shrinkage and Selection Operator regression were adopted in identifying significantly different infiltrating immune cells and differentially expressed genes (DEGs). Functional enrichment analysis was performed by Kyoto Encyclopedia of Genes and Genomes and Gene Ontology. The correlation between genes and immune cells was evaluated via Spearman's correlation analysis. Receiver Operator Characteristic curve analysis evaluated the markers' diagnostic effectiveness. The mRNA-miRNA and Drug-Gene-Immune cell interaction networks were constructed to identify potential diagnostic and therapeutic targets.

RESULTS

Plasma cells, M1 macrophages, M2 macrophages, neutrophils, eosinophils, and activated NK cells were the significantly different infiltrating immune cells in meningioma. A total of 951 DEGs, associated with synaptic function and structure, ion transport regulation, brain function, and immune-related pathways, were identified. Among 11 hub DEGs, RYR2 and TTR were correlated with plasma cells; SNCG was associated with NK cells; ADCY1 exhibited excellent diagnostic effectiveness; and ADCY1, BMX, KCNA5, SLCO4A1, and TTR could be considered as therapeutic targets.

CONCLUSIONS

ADCY1 can be identified as a diagnostic marker; ADCY1, BMX, KCNA5, SLCO4A1, and TTR are potential therapeutic targets, and their associations with macrophages, neutrophils, NK cells, and plasma cells might impact the tumorigenesis of meningiomas.

A machine learning model for colorectal liver metastasis post-hepatectomy prognostications

Background

Currently, surgical resection is the mainstay for colorectal liver metastases (CRLM) management and the only potentially curative treatment modality. Prognostication tools can support patient selection for surgical resection to maximize therapeutic benefit. This study aimed to develop a survival prediction model using machine learning based on a multicenter patient sample in Hong Kong.

Methods

Patients who underwent hepatectomy for CRLM between 1 January 2009 and 31 December 2018 in four hospitals in Hong Kong were included in the study. Survival analysis was performed using Cox proportional hazards (CPH). A stepwise selection on Cox multivariable models with Least Absolute Shrinkage and Selection Operator (LASSO) regression was applied to a multiply-imputed dataset to build a prediction model. The model was validated in the validation set, and its performance was compared with that of Fong Clinical Risk Score (CRS) using concordance index.

Results

A total of 572 patients were included with a median follow-up of 3.6 years. The full models for overall survival (OS) and recurrence-free survival (RFS) consist of the same 8 established and novel variables, namely colorectal cancer nodal stage, CRLM neoadjuvant treatment, Charlson Comorbidity Score, pre-hepatectomy bilirubin and carcinoembryonic antigen (CEA) levels, CRLM largest tumor diameter, extrahepatic metastasis detected on positron emission-tomography (PET)-scan as well as KRAS status. Our CRLM Machine-learning Algorithm Prognostication model (CMAP) demonstrated better ability to predict OS (C-index =0.651), compared with the Fong CRS for 1-year (C-index =0.571) and 5-year OS (C-index =0.574). It also achieved a C-index of 0.651 for RFS.

Conclusions

We present a promising machine learning algorithm to individualize prognostications for patients following resection of CRLM with good discriminative ability.

Intersectional characterization of emergency department (ED) staff experiences of racism: a survey of ED healthcare workers for the Disrupting Racism in Emergency Medicine (DRiEM) Investigators

INTRODUCTION

The impact of racism on patient outcomes in Emergency Medicine has been examined but there have been few studies exploring the experiences of racism in health care workers. This survey aims to explore the experience of racism by interdisciplinary staff in a tertiary ED. By characterizing the staff experience of racism in the ED, we hope to inform the design of strategies to disrupt racism and ultimately improve the health and wellness of both staff and patients.

METHODS

We conducted a self-administered, cross-sectional survey to explore the reported experience of racism by healthcare workers in a single urban ED in an academic trauma centre. We employed classification and regression tree analyses to evaluate predictors of racism through an intersectional lens.

RESULTS

A majority (n = 200, 75%) of all ED staff reported experiencing interpersonal racism (including physical violence, direct verbal violence, mistreatment and/or microaggressions) in the workplace. Respondents who identified as racialized self-reported significantly more racism at work than white respondents (86% vs. 63%, p < 0.001). Occupation, race, migrant status and age were identified through intersectional machine-learning models to be significantly predictive of the experience of racism. Nearly all respondents felt that the disruption of racism in Emergency medicine is important to them (90%, n = 207) and (93%, n = 214) were willing to participate in further training in anti-racism.

CONCLUSIONS

Racism against interdisciplinary staff working in EDs is common and the burden on healthcare workers is high. Intersections of occupation, race, age and migrant status are uniquely predictive of the experience of racism for EM staff. Interventions to disrupt racism should be informed by intersectional considerations to create a safe working environment and target populations most at risk. ED healthcare workers are willing to take steps to disrupt racism in their workplace and need institutional support to do so.

18F‑FDG PET/CT based radiomics features improve prediction of prognosis: multiple machine learning algorithms and multimodality applications for multiple myeloma

PURPOSE

Multiple myeloma (MM), the second most hematological malignancy, have been studied extensively in the prognosis of the clinical parameters, however there are only a few studies have discussed the role of dual modalities and multiple algorithms of ¹⁸F-FDG (¹⁸F-fluorodeoxyglucose) PET/CT based radiomics signatures for prognosis in MM patients. We hope to deeply mine the utility of raiomics data in the prognosis of MM.

METHODS

We extensively explored the predictive ability and clinical decision-making ability of different combination image data of PET, CT, clinical parameters and six machine learning algorithms, Cox proportional hazards model (Cox), linear gradient boosting models based on Cox's partial likelihood (GB-Cox), Cox model by likelihood based boosting (CoxBoost), generalized boosted regression modelling (GBM), random forests for survival model (RFS) and support vector regression for censored data model (SVCR). And the model evaluation methods include Harrell concordance index, time dependent receiver operating characteristic (ROC) curve, and decision curve analysis (DCA).

RESULTS

We finally confirmed 5 PET based features, and 4 CT based features, as well as 6 clinical derived features significantly related to progression free survival (PFS) and we included them in the model construction. In various modalities combinations, RSF and GBM algorithms significantly improved the accuracy and clinical net benefit of predicting prognosis compared with other algorithms. For all combinations of various modalities based models, single-modality PET based prognostic models' performance was outperformed baseline clinical parameters based models, while the performance of models of PET and CT combined with clinical parameters was significantly improved in various algorithms.

CONCLUSION

¹⁸F‑FDG PET/CT based radiomics models implemented with machine learning algorithms can significantly improve the clinical prediction of progress and increased clinical benefits providing prospects for clinical prognostic stratification for precision treatment as well as new research areas.

The effect of residential greenness during pregnancy on infant neurodevelopment using propensity score weighting: A prospective mother-infant paired cohort study

BACKGROUND

While prior studies have suggested an association between green spaces and infant neurodevelopment, the causal effect of green space exposure during pregnancy has not been fully investigated. This study aimed to identify with causal inference the effect of exposure to residential greenness during pregnancy on infants' mental-psychomotor development and the role of maternal education in modifying this association.

METHODS

We prospectively collected data of pregnant women and their infants from Mothers and Children Environmental Health cohort study. Based on residential addresses, we compiled information on the percent of green space using different buffer distances (100 m, 300 m, and 500 m) and air pollution (PM_2.5). Infant neurodevelopment was measured at 6 months of age using the Korean Bayley Scales of Infant Development II Mental Developmental Index (MDI) and Psychomotor Developmental Index (PDI). Generalized propensity scores (GPSs) were estimated from machine-learning (ML) algorithms. We deduced causal inference through GPS adjustment and weighting approaches. Further analyses confirmed whether the association was altered by maternal academic background.

RESULTS

A total of 845 mother-infant pairs from the cohort study were included. We found that exposure to green spaces was robustly associated with infants' mental development. For example, an increase in % green space within 300 m increased the MDI by 14.32 (95 % confidence interval [CI], 3.44-25.2) in the weighting approach. Additionally, the association was even more noticeable for mothers with college degrees or above: an increase in % green space within 300 m increased the MDI by 23.69 (95 % CI, 8.53-38.85) and the PDI by 22.45 (95 % CI, 2.58-42.33) in the weighting approach. This association did not appear in mothers without college degrees.

CONCLUSION

Exposure to green spaces during pregnancy showed a beneficial relationship with infant mental development. Maternal academic background could modify the impact of green space exposure on infant neurodevelopment.

Genome-wide associations, polygenic risk, and Mendelian randomization reveal limited interactions between John Henryism and cynicism

BACKGROUND

John Henryism (JH) is a strategy for dealing with chronic psychological stress characterized by high levels of physical effort and work. Cynicism is a belief that people are motivated primarily by self-interest. High scores on the JH scale and cynicism measures correlate with an increased risk of cardiovascular disease. High cynicism is also a hallmark of burnout syndrome, another known risk factor for heart disease.

AIM

To evaluate possible interactions between JH and cynicism hoping to clarify risk factors of burnout.

METHODS

We analyzed genetic and psychological data available from the Database of Genotypes and Phenotypes for genome-wide associations with these traits. We split the total available samples and used plink to perform the association studies on the discovery set (n = 1852, 80%) and tested for replication using the validation set (n = 465). We used scikit-learn to perform supervised machine learning for developing genetic risk algorithms.

RESULTS

We identified 2, 727, and 204 genetic associations for scores on the JH, cynicism and cynical distrust (CD) scales, respectively. We also found 173 associations with high cynicism, 109 with high CD, but no associations with high JH. We also produced polygenic classifiers for high cynicism using machine learning with areas under the receiver operator characteristics curve greater than 0.7.

CONCLUSION

We found significant genetic components to these traits but no evidence of an interaction. Therefore, while there may be a genetic risk, JH is not likely a burnout risk factor.

Estimation of reduced greenhouse gas emission from municipal solid waste incineration with electricity recovery in prefecture- and county-level cities of China

Municipal solid waste (MSW) without proper managements could be a significant source of greenhouse gas (GHG) emissions. MSW incineration with electricity recovery (MSW-IER) is recognized as a sustainable way to utilize waste, but its effectiveness on reducing GHG emissions at the city scale in China remain unclear due to limited data of MSW compositions. The aim of the study is to investigate reduction potential of GHG from MSW-IER in China. Based on the MSW compositions covering 106 Chinese prefecture-level cities during the period of 1985 to 2016, random forest models were built to predict MSW compositions in Chinese cities. MSW compositions in 297 cities of China from 2002 to 2017 were predicted using the model trained by a combination of socio-economic, climate and spatiotemporal factors. Spatiotemporal and climatic factors (such as economic development level, precipitation) accounted for 6.5 %-20.7 % and 20.1 %-37.6 % to total contributions on MSW composition, respectively. The GHG emissions from MSW-IER in each Chinese city were further calculated based on the predicted MSW compositions. The plastic is the main GHG emission source, accounting for over 91 % of the total emission during 2002-2017. Compared to baseline (landfill) emission, the GHG emission reduction from MSW-IER was 12.5 × 10⁷ kg CO₂-eq in 2002 and 415 × 10⁷ kg CO₂-eq in 2017, with an average annual growth rate of 26.3 %. The results provide basic data for estimating GHG emission in MSW management in China.

DePolymerase Predictor (DePP): a machine learning tool for the targeted identification of phage depolymerases

Biofilm production plays a clinically significant role in the pathogenicity of many bacteria, limiting our ability to apply antimicrobial agents and contributing in particular to the pathogenesis of chronic infections. Bacteriophage depolymerases, leveraged by these viruses to circumvent biofilm mediated resistance, represent a potentially powerful weapon in the fight against antibiotic resistant bacteria. Such enzymes are able to degrade the extracellular matrix that is integral to the formation of all biofilms and as such would allow complementary therapies or disinfection procedures to be successfully applied. In this manuscript, we describe the development and application of a machine learning based approach towards the identification of phage depolymerases. We demonstrate that on the basis of a relatively limited number of experimentally proven enzymes and using an amino acid derived feature vector that the development of a powerful model with an accuracy on the order of 90% is possible, showing the value of such approaches in protein functional annotation and the discovery of novel therapeutic agents.

Gut microbiome combined with metabolomics reveals biomarkers and pathways in central precocious puberty

BACKGROUND

Central precocious puberty (CPP) is a common disease in prepubertal children and results mainly from disorders in the endocrine system. Emerging evidence has highlighted the involvement of gut microbes in hormone secretion, but their roles and downstream metabolic pathways in CPP remain unknown.

METHODS

To explore the gut microbes and metabolism alterations in CPP, we performed the 16S rRNA sequencing and untargeted metabolomics profiling for 91 CPP patients and 59 healthy controls. Bioinformatics and statistical analyses, including the comparisons of alpha and beta diversity, abundances of microbes, were undertaken on the 16S rRNA gene sequences and metabolism profiling. Classifiers were constructed based on the microorganisms and metabolites. Functional and pathway enrichment analyses were performed for identification of the altered microorganisms and metabolites in CPP.

RESULTS

We integrated a multi-omics approach to investigate the alterations and functional characteristics of gut microbes and metabolites in CPP patients. The fecal microbiome profiles and fecal and blood metabolite profiles for 91 CPP patients and 59 healthy controls were generated and compared. We identified the altered microorganisms and metabolites during the development of CPP and constructed a machine learning-based classifier for distinguishing CPP. The Area Under Curves (AUCs) of the classifies were ranged from 0.832 to 1.00. In addition, functional analysis of the gut microbiota revealed that the nitric oxide synthesis was closely associated with the progression of CPP. Finally, we investigated the metabolic potential of gut microbes and discovered the genus Streptococcus could be a candidate molecular marker for CPP treatment.

CONCLUSIONS

Overall, we utilized multi-omics data from microorganisms and metabolites to build a classifier for discriminating CPP patients from the common populations and recognized potential therapeutic molecular markers for CPP through comprehensive analyses.

Accurate and non-destructive monitoring of mold contamination in foodstuffs based on whole-cell biosensor array coupling with machine-learning prediction models

Mold contamination in foodstuffs causes huge economic losses, quality deterioration and mycotoxin production. Thus, non-destructive and accurate monitoring of mold occurrence in foodstuffs is highly required. We proposed a novel whole-cell biosensor array to monitor pre-mold events in foodstuffs. Firstly, 3 volatile markers ethyl propionate, 1-methyl-1 H-pyrrole and 2,3-butanediol were identified from pre-mold peanuts using gas chromatography-mass spectrometry. Together with other 3 frequently-reported volatiles from Aspergillus flavus infection, the volatiles at subinhibitory concentrations induced significant but differential response patterns from 14 stress-responsive Escherichia coli promoters. Subsequently, a whole-cell biosensor array based on the 14 promoters was constructed after whole-cell immobilization in calcium alginate. To discriminate the response patterns of the whole-cell biosensor array to mold-contaminated foodstuffs, optimal classifiers were determined by comparing 6 machine-learning algorithms. 100 % accuracy was achieved to discriminate healthy from moldy peanuts and maize, and 95 % and 98 % accuracy in discriminating pre-mold stages for infected peanuts and maize, based on random forest classifiers. 83 % accuracy was obtained to separate moldy peanuts from moldy maize by sparse partial least square determination analysis. The results demonstrated high accuracy and practicality of our method based on a whole-cell biosensor array coupling with machine-learning classifiers for mold monitoring in foodstuffs.