Harnessing Deep Learning Methods for Voltage-Gated Ion Channel Drug Discovery.

Voltage-gated ion channels (VGICs) are pivotal in regulating electrical activity in excitable cells and are critical pharmaceutical targets for treating many diseases including cardiac arrhythmia and neuropathic pain. Despite their significance, challenges such as achieving target selectivity persist in VGIC drug development. Recent progress in deep learning, particularly diffusion models, has enabled the computational design of protein binders for any clinically relevant protein based solely on its structure. These developments coincide with a surge in experimental structural data for VGICs, providing a rich foundation for computational design efforts. This review explores the recent advancements in computational protein design using deep learning and diffusion methods, focusing on their application in designing protein binders to modulate VGIC activity. We discuss the potential use of these methods to computationally design protein binders targeting different regions of VGICs, including the pore domain, voltage-sensing domains, and interface with auxiliary subunits. We provide a comprehensive overview of the different design scenarios, discuss key structural considerations, and address the practical challenges in developing VGIC-targeting protein binders. By exploring these innovative computational methods, we aim to provide a framework for developing novel strategies that could significantly advance VGIC pharmacology and lead to the discovery of effective and safe therapeutics.

Evaluation of penalized and machine learning methods for asthma disease prediction in the Korean Genome and Epidemiology Study (KoGES).

BACKGROUND: Genome-wide association studies have successfully identified genetic variants associated with human disease. Various statistical approaches based on penalized and machine learning methods have recently been proposed for disease prediction. In this study, we evaluated the performance of several such methods for predicting asthma using the Korean Chip (KORV1.1) from the Korean Genome and Epidemiology Study (KoGES). RESULTS: First, single-nucleotide polymorphisms were selected via single-variant tests using logistic regression with the adjustment of several epidemiological factors. Next, we evaluated the following methods for disease prediction: ridge, least absolute shrinkage and selection operator, elastic net, smoothly clipped absolute deviation, support vector machine, random forest, boosting, bagging, naïve Bayes, and k-nearest neighbor. Finally, we compared their predictive performance based on the area under the curve of the receiver operating characteristic curves, precision, recall, F1-score, Cohen's Kappa, balanced accuracy, error rate, Matthews correlation coefficient, and area under the precision-recall curve. Additionally, three oversampling algorithms are used to deal with imbalance problems. CONCLUSIONS: Our results show that penalized methods exhibit better predictive performance for asthma than that achieved via machine learning methods. On the other hand, in the oversampling study, randomforest and boosting methods overall showed better prediction performance than penalized methods.

Genomic prediction using machine learning: a comparison of the performance of regularized regression, ensemble, instance-based and deep learning methods on synthetic and empirical data.

BACKGROUND: The accurate prediction of genomic breeding values is central to genomic selection in both plant and animal breeding studies. Genomic prediction involves the use of thousands of molecular markers spanning the entire genome and therefore requires methods able to efficiently handle high dimensional data. Not surprisingly, machine learning methods are becoming widely advocated for and used in genomic prediction studies. These methods encompass different groups of supervised and unsupervised learning methods. Although several studies have compared the predictive performances of individual methods, studies comparing the predictive performance of different groups of methods are rare. However, such studies are crucial for identifying (i) groups of methods with superior genomic predictive performance and assessing (ii) the merits and demerits of such groups of methods relative to each other and to the established classical methods. Here, we comparatively evaluate the genomic predictive performance and informally assess the computational cost of several groups of supervised machine learning methods, specifically, regularized regression methods, deep, ensemble and instance-based learning algorithms, using one simulated animal breeding dataset and three empirical maize breeding datasets obtained from a commercial breeding program. RESULTS: Our results show that the relative predictive performance and computational expense of the groups of machine learning methods depend upon both the data and target traits and that for classical regularized methods, increasing model complexity can incur huge computational costs but does not necessarily always improve predictive accuracy. Thus, despite their greater complexity and computational burden, neither the adaptive nor the group regularized methods clearly improved upon the results of their simple regularized counterparts. This rules out selection of one procedure among machine learning methods for routine use in genomic prediction. The results also show that, because of their competitive predictive performance, computational efficiency, simplicity and therefore relatively few tuning parameters, the classical linear mixed model and regularized regression methods are likely to remain strong contenders for genomic prediction. CONCLUSIONS: The dependence of predictive performance and computational burden on target datasets and traits call for increasing investments in enhancing the computational efficiency of machine learning algorithms and computing resources.

Predicting redox potentials by graph-based machine learning methods.

The evaluation of oxidation and reduction potentials is a pivotal task in various chemical fields. However, their accurate prediction by theoretical computations, which is a complementary task and sometimes the only alternative to experimental measurement, may be often resource-intensive and time-consuming. This paper addresses this challenge through the application of machine learning techniques, with a particular focus on graph-based methods (such as graph edit distances, graph kernels, and graph neural networks) that are reviewed to enlighten their deep links with theoretical chemistry. To this aim, we establish the ORedOx159 database, a comprehensive, homogeneous (with reference values stemming from density functional theory calculations), and reliable resource containing 318 one-electron reduction and oxidation reactions and featuring 159 large organic compounds. Subsequently, we provide an instructive overview of the good practice in machine learning and of commonly utilized machine learning models. We then assess their predictive performances on the ORedOx159 dataset through extensive analyses. Our simulations using descriptors that are computed in an almost instantaneous way result in a notable improvement in prediction accuracy, with mean absolute error (MAE) values equal to 5.6 kcal mol - 1 for reduction and 7.2 kcal mol - 1 for oxidation potentials, which paves a way toward efficient in silico design of new electrochemical systems.

Investigating mental healthcare inequalities associated with forced migration: promise and potential pitfalls of electronic health records.

There is a lack of data on mental health service utilisation and outcomes for people with experience of forced migration living in the UK. Details about migration experiences documented in free-text fields in electronic health records might be harnessed using novel data science methods; however, there are potential limitations and ethical concerns.

Global assessment of the knowledge and confidence in managing allergic disorders among primary care pediatricians across Europe: An EAACI task force report.

BACKGROUND: Pediatricians are often the first point of contact for children in Primary Care (PC), but still perceive gaps in their allergy knowledge. We investigated self-perceived knowledge gaps and educational needs in pediatricians across healthcare systems in Europe so that future educational initiatives may better support the delivery of allergy services in PC. METHOD: A multinational survey was circulated to pediatricians who care for children and adolescents with allergy problems in PC by the EAACI Allergy Educational Needs in Primary Care Pediatricians Task Force from February to March 2023. A 5-point Likert scale was used to assess the level of agreement with questionnaire statements. Thirty surveys per country were the cut-off for inclusion and statistical analysis. RESULTS: In this study, 1991 respondents were obtained from 56 countries across Europe and 210 responses were from countries with a cut-off below 30 participants per country. Primary care pediatricians (PCPs) comprised 74.4% of the respondents. The majority (65.3%) were contracted to state or district health services. 61.7% had awareness of guidelines for onward allergy referral in their countries but only 22.3% were aware of the EAACI competencies document for allied health professionals for allergy. Total sample respondents versus PCPs showed 52% and 47% of them have access to allergy investigations in their PC facility (mainly specific IgE and skin prick tests); 67.6% and 58.9% have access to immunotherapy, respectively. The main barrier to referral to a specialist was a consideration that the patient's condition could be diagnosed and treated in this PC facility, (57.8% and 63.6% respectively). The main reasons for referral were the need for hospital assessment, and partial response to first-line treatment (55.4% and 59.2%, 47% and 50.7%, respectively). Learning and assessment methods preference was fairly equally divided between Traditional methods (45.7% and 50.1% respectively) and e-learning 45.5% and 44.9%, respectively. Generalist physicians (GPs) have the poorest access to allergy investigations (32.7%, p = .000). The majority of the total sample (91.9%) assess patients with allergic pathology. 868 (43.6%) and 1117 (46.1%), received allergy training as undergraduates and postgraduates respectively [these proportions in PCPs were higher (45% and 59%), respectively]. PCPs with a special interest in allergology experienced greater exposure to allergy teaching as postgraduates. GPs received the largest amount of allergy teaching as undergraduates. Identifying allergic disease based on clinical presentation, respondents felt most confident in the management of eczema/atopic dermatitis (87.4%) and rhinitis/asthma (86.2%), and least confident in allergen immunotherapy (36.9%) and latex allergy (30.8%). CONCLUSION: This study exploring the confidence of PCPs to diagnose, manage, and refer patients with allergies, demonstrated knowledge gaps and educational needs for allergy clinical practice. It detects areas in need of urgent improvement especially in latex and allergen immunotherapy. It is important to ensure the dissemination of allergy guidelines and supporting EAACI documents since the majority of PCPs lack awareness of them. This survey has enabled us to identify what the educational priorities of PCPs are and how they would like to have them met.

Forecast of agri-residues generation from rice, wheat and oilseed crops in India using machine learning techniques: Exploring strategies for sustainable smart management.

Management of agri-residues generated in large quantities necessitates for its accurate estimation. Data analysis using machine learning methods can predict the agri-residues generation. The objective of the study was to forecast agri-residues generation from rice, wheat, and oilseed crops in India using ML methods and their sustainable uses. Prediction of agri-residues was done first by forecasting the crop production via the application of ML techniques for the period 2022 to 2030, and then the amount of crop residues generation calculated by multiplying the crop productions with the residues-to-product-ratio (RPR) values of the respective crops. RPR was estimated by using the gravimetric ratio of the residue to the actual crop production. The crop-specific RPR values were taken from various earlier studies in Indian context. The RPR values of 1.73 for the rice, 1.65 for wheat, and 2.6 for the oilseed crop were used as a conversion factor for residues calculation. Machine learning models linear regression, sequential minimal optimization regression (SMOreg), M5 Rule, and Gaussian process were used in the study. SMOreg performed better in models tested by coefficient of determination, root mean square error, and mean absolute error. The models predicted the generation of residues in 2030 as rice straw and husk 195.76 Mt to 277.68 Mt, wheat straw 188.62 Mt to 266.95 Mt, and oilseed stalk and oil cakes 55.61 Mt to 96.30 Mt in India. An overview of the management of agri-residues discussed. Estimation of agri-residues can provide an opportunity to utilize them with the best possible ways, lessen pollution and promote a zero-waste strategy.

Predicting academic achievement from the collaborative influences of executive function, physical fitness, and demographic factors among primary school students in China: ensemble learning methods.

BACKGROUND: Elevated levels of executive function and physical fitness play a pivotal role in shaping future quality of life. However, few studies have examined the collaborative influences of physical and mental health on academic achievement. This study aims to investigate the key factors that collaboratively influence primary school students' academic achievement from executive function, physical fitness, and demographic factors. Additionally, ensemble learning methods are employed to predict academic achievement, and their predictive performance is compared with individual learners. METHODS: A cluster sampling method was utilized to select 353 primary school students from Huai'an, China, who underwent assessments for executive function, physical fitness, and academic achievement. The recursive feature elimination cross-validation method was employed to identify key factors that collaboratively influence academic achievement. Ensemble learning models, utilizing eXtreme Gradient Boosting and Random Forest algorithms, were constructed based on Bagging and Boosting methods. Individual learners were developed using Support Vector Machine, Decision Tree, Logistic Regression, and Linear Discriminant Analysis algorithms, followed by the establishment of a Stacking ensemble learning model. RESULTS: Our findings revealed that sex, body mass index, muscle strength, cardiorespiratory function, inhibition, working memory, and shifting were key factors influencing the academic achievement of primary school students. Moreover, ensemble learning models demonstrated superior predictive performance compared to individual learners in predicting academic achievement among primary school students. CONCLUSIONS: Our results suggest that recognizing sex differences and emphasizing the simultaneous development of cognition and physical well-being can positively impact the academic development of primary school students. Ensemble learning methods warrant further attention, as they enable the establishment of an accurate academic early warning system for primary school students.

Evaluating the Association of Anthropometric Indices With Total Cholesterol in a Large Population Using Data Mining Algorithms.

BACKGROUND: Unbalanced levels of serum total cholesterol (TC) and its subgroups are called dyslipidemia. Several anthropometric indices have been developed to provide a more accurate assessment of body shape and the health risks associated with obesity. In this study, we used the random forest model (RF), decision tree (DT), and logistic regression (LR) to predict total cholesterol based on new anthropometric indices in a sex-stratified analysis. METHOD: Our sample size was 9639 people in which anthropometric parameters were measured for the participants and data regarding the demographic and laboratory data were obtained. Aiding the machine learning, DT, LR, and RF were drawn to build a measurement prediction model. RESULTS: Anthropometric and other related variables were compared between both TC <200 and TC ≥200 groups. In both males and females, Lipid Accumulation Product (LAP) had the greatest effect on the risk of TC increase. According to results of the RF model, LAP and Visceral Adiposity Index (VAI) were significant variables for men. VAI also had a stronger correlation with HDL-C and triglyceride. We identified specific anthropometric thresholds based on DT analysis that could be used to classify individuals at high or low risk of elevated TC levels. The RF model determined that the most important variables for both genders were VAI and LAP. CONCLUSION: We tend to present a picture of the Persian population's anthropometric factors and their association with TC level and possible risk factors. Various anthropometric indices indicated different predictive power for TC levels in the Persian population.

Multiple machine-learning tools identifying prognostic biomarkers for acute Myeloid Leukemia.

BACKGROUND: Acute Myeloid Leukemia (AML) generally has a relatively low survival rate after treatment. There is an urgent need to find new biomarkers that may improve the survival prognosis of patients. Machine-learning tools are more and more widely used in the screening of biomarkers. METHODS: Least Absolute Shrinkage and Selection Operator (LASSO), Support Vector Machine-Recursive Feature Elimination (SVM-RFE), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), lrFuncs, IdaProfile, caretFuncs, and nbFuncs models were used to screen key genes closely associated with AML. Then, based on the Cancer Genome Atlas (TCGA), pan-cancer analysis was performed to determine the correlation between important genes and AML or other cancers. Finally, the diagnostic value of important genes for AML was verified in different data sets. RESULTS: The survival analysis results of the training set showed 26 genes with survival differences. After the intersection of the results of each machine learning method, DNM1, MEIS1, and SUSD3 were selected as key genes for subsequent analysis. The results of the pan-cancer analysis showed that MEIS1 and DNM1 were significantly highly expressed in AML; MEIS1 and SUSD3 are potential risk factors for the prognosis of AML, and DNM1 is a potential protective factor. Three key genes were significantly associated with AML immune subtypes and multiple immune checkpoints in AML. The results of the verification analysis show that DNM1, MEIS1, and SUSD3 have potential diagnostic value for AML. CONCLUSION: Multiple machine learning methods identified DNM1, MEIS1, and SUSD3 can be regarded as prognostic biomarkers for AML.

Comparative analysis of different Karnal bunt disease prediction models developed by machine learning techniques for Punjab conditions.

Timely prediction of pathogen is important key factor to reduce the quality and yield losses. Wheat is major crop in northern part of India. In Punjab, wheat face challenge by different diseases so the study was conducted for two locations viz. Ludhiana and Bathinda. The information regarding the occurrence of Karnal bunt in 12 consecutive crop seasons (from 2009-10 to 2020-21) in Ludhiana district and in 9 crop seasons (from 2010-11 to 2018-19) in Bathinda district, was collected from the Wheat Section of the Department of Plant Breeding and Genetics at Punjab Agricultural University (PAU), located in Ludhiana. The study aims to investigate the adequacy of various methods of machine learning for prediction of Karnal bunt using meteorological data for different time period viz. February, March, 15 February to 15 March and overall period obtained from Department of Climate Change and Agricultural Meteorology, PAU, Ludhiana. The most intriguing outcome is that for each period, different disease prediction models performed well. The random forest regression (RF) for February month, support vector regression (SVR) for March month, SVR and BLASSO for 15 February to 15 March period and random forest for overall period surpassed the performance than other models. The Taylor diagram was created to assess the effectiveness of intricate models by comparing various metrics such as root mean square error (RMSE), root relative square error (RRSE), correlation coefficient (r), relative mean absolute error (MAE), modified D-index, and modified NSE. It allows for a comprehensive evaluation of these models' performance.

Structure and quality of bedside teaching: A videographic analysis.

Bedside teaching (BST) is a core element of medical education. In light of a reported decrease in BST, evidence on how to use BST time most efficiently should be developed. Given that little research into the tangible quality characteristics of good BST has been reported, we investigated the influence of various teacher and structural characteristics on the perceived quality of BST. We filmed and coded 36 BSTs involving 24 lecturers and 259 students. Structural characteristics of interest were: number of students and patients, overall duration, and the proportion of clinical examination. Lecturer questionnaires provided insight into teaching experience and intrinsic motivation, and student evaluations assessed the quality of BSTs in three dimensions. Correlations are reported using the Pearson r and a linear mixed model (LMM). The intrinsic motivation of lecturers was significantly positively correlated with perceived quality, but their experience was only weakly so correlated. In the LMM, a significant association was observed for the quality dimension of clinical teaching with the number of patients and the proportion of time spent on clinical examination. Based on our findings, we recommend including multiple patients in BSTs, and providing substantial opportunities for clinical examination. Regarding lecturers, motivation matters more than experience.

Game-based learning enhances students' understanding of endocrine physiology in veterinary medicine.

Endocrine physiology is a complex subject for students. Game-based learning (GBL) and case-based learning (CBL) are active methodologies that are widely used because of their potential for motivation and greater proximity to the reality of modern students. We evaluated the effectiveness of GBL and CBL among veterinary medicine students compared with a control group using peer tutoring. Students (n = 106) from two institutions volunteered to participate in this study. The participants were submitted to a pretest questionnaire and subsequently were divided into three paired groups by their performance on the pretest exam: 1) traditional class + peer tutoring, 2) traditional class + GBL, and 3) traditional class + CBL. After the students completed the activities, their performance was once again evaluated by applying a new test with the same initial 10 questions and another set of 10 different questions. The students' perceptions and satisfaction with the methodologies and learning strategies were assessed. Anxiety was assessed with the State-Trait Anxiety Inventory before and after the conventional class and after the active methodologies. The GBL group significantly improved their correct answers compared with the baseline (P < 0.05), with no significant difference from CBL and peer tutoring. Anxiety levels did not differ regardless of the time of evaluation or the teaching methodology applied. GBL promoted a greater perception of the stimulus for self-study and problem-solving ability and contributed to the development of group dynamics compared with the group who received CBL (P < 0.05). In conclusion, GBL showed better results than peer tutoring and CBL.NEW & NOTEWORTHY We compared the supplementary use of game-based learning, case-based learning, and peer tutoring in the study of endocrine physiology by veterinary students and observed a slight advantage for game-based learning over the other two methodologies. The game was developed by the authors and is an unprecedented tool that can prove useful to improve knowledge acquisition in students of veterinary medicine. Thus, game-based learning is an effective supplementary teaching strategy.

Exploring dental and oral hygiene students' interprofessional readiness: a cross-sectional study in joint paediatric outreach training.

BACKGROUND: Interprofessional education is vital in oral healthcare education and should be integrated into both theoretical and work-based education. Little research addresses interprofessional education in dental hands-on training in authentic oral healthcare settings. The aim of the study was to examine the readiness and attitudes of dental and oral hygiene students towards interprofessional education during joint paediatric outreach training. METHODS: In the spring of 2022, a cross-sectional study was done involving dental and oral hygiene students using the Readiness for Interprofessional Learning Scale (RIPLS) during joint paediatric outreach training. The 19-item tool was answered on a five-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree). Means, standard deviations, minimums, maximums, and medians were calculated for each subscale and overall score. Students grouped according to their categorical variables were compared for statistically significant differences. The Mann-Whitney U-test was used for groups of two and the Kruskal-Wallis one-way analysis for groups of three or more. The internal consistency of the scale was measured with Cronbach's alpha. Statistical level was set at 0.05. RESULTS: The survey included 111 participants, consisting of 51 oral hygiene students and 60 dental students, with a response rate of 93%. The questionnaire yielded a high overall mean score of 4.2. Both oral hygiene (4.3) and dental students (4.2) displayed strong readiness for interprofessional education measured by the RIPLS. The subscale of teamwork and collaboration achieved the highest score of 4.5. Students lacking prior healthcare education or work experience obtained higher RIPLS scores. Oral hygiene students rated overall items (p = 0.019) and the subscales of positive professional identity (p = < 0.001) and roles and responsibilities (p = 0.038) higher than dental students. The Cronbach's alpha represented high internal consistency for overall RIPLS scores on the scale (0.812). CONCLUSIONS: Both oral hygiene and dental students perceived shared learning as beneficial and showcased high readiness for interprofessional education, as evident in their RIPLS scores. Integrating interprofessional learning into oral hygiene and dental curricula is important. Studying together can form a good basis for future working life collaboration.

A survey on adverse drug reaction studies: data, tasks and machine learning methods.

MOTIVATION: Adverse drug reaction (ADR) or drug side effect studies play a crucial role in drug discovery. Recently, with the rapid increase of both clinical and non-clinical data, machine learning methods have emerged as prominent tools to support analyzing and predicting ADRs. Nonetheless, there are still remaining challenges in ADR studies. RESULTS: In this paper, we summarized ADR data sources and review ADR studies in three tasks: drug-ADR benchmark data creation, drug-ADR prediction and ADR mechanism analysis. We focused on machine learning methods used in each task and then compare performances of the methods on the drug-ADR prediction task. Finally, we discussed open problems for further ADR studies. AVAILABILITY: Data and code are available at https://github.com/anhnda/ADRPModels.

Evaluation of an automated genome interpretation model for rare disease routinely used in a clinical genetic laboratory.

PURPOSE: The analysis of exome and genome sequencing data for the diagnosis of rare diseases is challenging and time-consuming. In this study, we evaluated an artificial intelligence model, based on machine learning for automating variant prioritization for diagnosing rare genetic diseases in the Baylor Genetics clinical laboratory. METHODS: The automated analysis model was developed using a supervised learning approach based on thousands of manually curated variants. The model was evaluated on 2 cohorts. The model accuracy was determined using a retrospective cohort comprising 180 randomly selected exome cases (57 singletons, 123 trios); all of which were previously diagnosed and solved through manual interpretation. Diagnostic yield with the modified workflow was estimated using a prospective "production" cohort of 334 consecutive clinical cases. RESULTS: The model accurately pinpointed all manually reported variants as candidates. The reported variants were ranked in top 10 candidate variants in 98.4% (121/123) of trio cases, in 93.0% (53/57) of single proband cases, and 96.7% (174/180) of all cases. The accuracy of the model was reduced in some cases because of incomplete variant calling (eg, copy number variants) or incomplete phenotypic description. CONCLUSION: The automated model for case analysis assists clinical genetic laboratories in prioritizing candidate variants effectively. The use of such technology may facilitate the interpretation of genomic data for a large number of patients in the era of precision medicine.

Identification of ferroptosis-related genes in male mice with sepsis-induced acute lung injury based on transcriptome sequencing.

BACKGROUND: Sepsis can result in acute lung injury (ALI). Studies have shown that pharmacological inhibition of ferroptosis can treat ALI. However, the regulatory mechanisms of ferroptosis in sepsis-induced ALI remain unclear. METHODS: Transcriptome sequencing was performed on lung tissue samples from 10 sepsis-induced mouse models of ALI and 10 control mice. After quality control measures, clean data were used to screen for differentially expressed genes (DEGs) between the groups. The DEGs were then overlapped with ferroptosis-related genes (FRGs) to obtain ferroptosis-related DEGs (FR-DEGs). Subsequently, least absolute shrinkage and selection operator (Lasso) and Support Vector Machine-Recursive Feature Elimination (SVM-RFE) were used to obtain key genes. In addition, Ingenuity Pathway Analysis (IPA) was employed to explore the disease, function, and canonical pathways related to the key genes. Gene set enrichment analysis (GSEA) was used to investigate the functions of the key genes, and regulatory miRNAs of key genes were predicted using the NetworkAnalyst and StarBase databases. Finally, the expression of key genes was validated with the GSE165226 and GSE168796 datasets sourced from the Gene Expression Omnibus (GEO) database and using quantitative real-time polymerase chain reaction (qRT-PCR). RESULTS: Thirty-three FR-DEGs were identified between 1843 DEGs and 259 FRGs. Three key genes, Ncf2, Steap3, and Gclc, were identified based on diagnostic models established by the two machine learning methods. They are mainly involved in infection, immunity, and apoptosis, including lymphatic system cell migration and lymphocyte and T cell responses. Additionally, the GSEA suggested that Ncf2 and Steap3 were similarly enriched in mRNA processing, response to peptides, and leukocyte differentiation. Furthermore, a key gene-miRNA network including 2 key genes (Steap3 and Gclc) and 122 miRNAs, and a gene-miRNA network with 1 key gene (Steap3) and 3 miRNAs were constructed using NetworkAnalyst and StarBase, respectively. Both databases predicted that mmu-miR-15a-5p was the target miRNA of Steap3. Finally, Ncf2 expression was validated using both datasets and qRT-PCR, and Steap3 was validated using GSE165226 and qRT-PCR. CONCLUSIONS: This study identified two FR-DEGs (Ncf2 and Steap3) associated with sepsis-induced ALI via transcriptome analyses, as well as their functional and metabolic pathways.

Tuning Random Forests for Causal Inference under Cluster-Level Unmeasured Confounding.

Recently, there has been growing interest in using machine learning methods for causal inference due to their automatic and flexible ability to model the propensity score and the outcome model. However, almost all the machine learning methods for causal inference have been studied under the assumption of no unmeasured confounding and there is little work on handling omitted/unmeasured variable bias. This paper focuses on a machine learning method based on random forests known as Causal Forests and presents five simple modifications for tuning Causal Forests so that they are robust to cluster-level unmeasured confounding. Our simulation study finds that adjusting the default tuning procedure with the propensity score from fixed effects logistic regression or using variables that are centered to their cluster means produces estimates that are more robust to cluster-level unmeasured confounding. Also, when these parametric propensity score models are mis-specified, our modified machine learning methods remain robust to bias from cluster-level unmeasured confounders compared to existing parametric approaches based on propensity score weighting. We conclude by demonstrating our proposals in a real data study concerning the effect of taking an eighth-grade algebra course on math achievement scores from the Early Childhood Longitudinal Study.

Detection of False Data Injection Attacks in Smart Grids Based on Expectation Maximization.

The secure operation of smart grids is closely linked to state estimates that accurately reflect the physical characteristics of the grid. However, well-designed false data injection attacks (FDIAs) can manipulate the process of state estimation by injecting malicious data into the measurement data while bypassing the detection of the security system, ultimately causing the results of state estimation to deviate from secure values. Since FDIAs tampering with the measurement data of some buses will lead to error offset, this paper proposes an attack-detection algorithm based on statistical learning according to the different characteristic parameters of measurement error before and after tampering. In order to detect and classify false data from the measurement data, in this paper, we report the model establishment and estimation of error parameters for the tampered measurement data by combining the the k-means++ algorithm with the expectation maximization (EM) algorithm. At the same time, we located and recorded the bus that the attacker attempted to tamper with. In order to verify the feasibility of the algorithm proposed in this paper, the IEEE 5-bus standard test system and the IEEE 14-bus standard test system were used for simulation analysis. Numerical examples demonstrate that the combined use of the two algorithms can decrease the detection time to less than 0.011883 s and correctly locate the false data with a probability of more than 95%.

A Framework for Representing, Building and Reusing Novel State-of-the-Art Three-Dimensional Object Detection Models in Point Clouds Targeting Self-Driving Applications.

The rapid development of deep learning has brought novel methodologies for 3D object detection using LiDAR sensing technology. These improvements in precision and inference speed performances lead to notable high performance and real-time inference, which is especially important for self-driving purposes. However, the developments carried by these approaches overwhelm the research process in this area since new methods, technologies and software versions lead to different project necessities, specifications and requirements. Moreover, the improvements brought by the new methods may be due to improvements in newer versions of deep learning frameworks and not just the novelty and innovation of the model architecture. Thus, it has become crucial to create a framework with the same software versions, specifications and requirements that accommodate all these methodologies and allow for the easy introduction of new methods and models. A framework is proposed that abstracts the implementation, reusing and building of novel methods and models. The main idea is to facilitate the representation of state-of-the-art (SoA) approaches and simultaneously encourage the implementation of new approaches by reusing, improving and innovating modules in the proposed framework, which has the same software specifications to allow for a fair comparison. This makes it possible to determine if the key innovation approach outperforms the current SoA by comparing models in a framework with the same software specifications and requirements.