Least Absolute Shrinkage And Selection Operator Research Articles

Deep learning-based CT image for pulmonary nodule classification with intrathoracic fat: A multicenter study

Correctly distinguishing between benign and malignant pulmonary nodules can avoid unnecessary invasive procedures. However, existing models exhibit a deficiency in the study of fat regions. We aim to create a multimodal deep learning nomogram (DLN) using pulmonary nodule and intrathoracic fat (ITF) features to distinguish between benign and malignant nodules and further demonstrate that intrathoracic fat can enhance the performance of lung cancer prediction models. A segmentation methodology was devised for precisely delineating intranodular and perinodular region (IPN). The Swin Transformer was used for feature extraction. The least absolute shrinkage and selection operator (LASSO) regression was employed for feature selection. The nomogram integrating IPN and ITF signatures revealed better diagnostic accuracy than IPN signature alone (net reclassification improvement = 0.134, P<0.05, and integrated discrimination improvement = 0.078, P<0.05). In the internal test cohort, DLN obtained an area under the curve (AUC) of 0.913 (95 % CI: 0.870, 0.952). In the external test cohort, the AUC was 0.906 (95 % CI: 0.867, 0.938). There was no significant difference between the calibrated curve and perfectly calibrated curve in all cohorts (P>0.05). Decision curve analysis showed that the DLN was clinically useful. The DLN demonstrated substantial diagnostic value for distinguishing between benign and malignant pulmonary nodules. The incorporation of the ITF signature has enhanced the performance of DLN.

Plasma proteomics and carotid intima-media thickness in the UK biobank cohort

Background and aimsUltrasound derived carotid intima-media thickness (cIMT) is valuable for cardiovascular risk stratification. We assessed the relative importance of traditional atherosclerosis risk factors and plasma proteins in predicting cIMT measured nearly a decade later.MethodWe examined 6,136 UK Biobank participants with 1,461 proteins profiled using the proximity extension assay applied to their baseline blood draw who subsequently underwent a cIMT measurement. We implemented linear regression, stepwise Akaike Information Criterion-based, and the least absolute shrinkage and selection operator (LASSO) models to identify potential proteomic as well as non-proteomic predictors. We evaluated our model performance using the proportion variance explained (R2).ResultThe mean time from baseline assessment to cIMT measurement was 9.2 years. Age, blood pressure, and anthropometric related variables were the strongest predictors of cIMT with fat-free mass index of the truncal region being the strongest predictor among adiposity measurements. A LASSO model incorporating variables including age, assessment center, genetic risk factors, smoking, blood pressure, trunk fat-free mass index, apolipoprotein B, and Townsend deprivation index combined with 97 proteins achieved the highest R2 (0.308, 95% C.I. 0.274, 0.341). In contrast, models built with proteins alone or non-proteomic variables alone explained a notably lower R2 (0.261, 0.228–0.294 and 0.260, 0.226–0.293, respectively). Chromogranin b (CHGB), Cystatin-M/E (CST6), leptin (LEP), and prolargin (PRELP) were the proteins consistently selected across all models.ConclusionPlasma proteins add to the clinical and genetic risk factors in predicting a cIMT measurement. Our findings implicate blood pressure and extracellular matrix-related proteins in cIMT pathophysiology.

Screening and validating genes associated with cuproptosis in systemic lupus erythematosus by expression profiling combined with machine learning.

Cell death has long been a focal point in life sciences research, and recently, scientists have discovered a novel form of cell death induced by copper, termed cuproptosis. This paper aimed to identify genes associated with cuproptosis in Systemic Lupus Erythematosus (SLE) through machine learning, combined with single-cell sequencing (scRNA-seq), to screen and validate related genes. The analytical results were then experimentally verified. Two published microarray gene expression datasets (GSE65391 and GSE61635) from SLE and control peripheral blood samples were downloaded from the GEO database. The GSE65391 dataset was used as the training group, while the GSE61635 dataset served as the validation group. Differentially expressed genes (DEGs) from GSE65391 identified 12 differential genes. Nine diagnostic genes, considered potential biomarkers, were selected using the least absolute shrinkage and selection operator (LASSO) and support vector machine recursive feature elimination (SVM-RFE) analysis. The Receiver Operating Characteristic (ROC) curves for both the training and validation groups were used to calculate the Area Under the Curve (AUC) to assess discriminatory properties. CIBERSORT was used to assess the relationship between these diagnostic genes and a reference set of infiltrating immune cells. Single-cell RNA sequencing data (GSE162577) from SLE patients were also obtained from the GEO database and analyzed. Experimental validation of the most important SLE biomarkers was performed. Twelve significantly different cuproptosis-related genes were identified in the GSE65391 training set. Immune cell analysis revealed 12 immune cell types and identified nine signature genes, including PDHB, GLS, DLAT, LIAS, MTF1, DLST, DLD, LIPT1, and FDX1. In the GSE61635 validation set, seven genes were weakly expressed, and two genes were strongly expressed in the treatment group. According to the ROC curves, PDHB and GLS demonstrated significant diagnostic value. Additionally, correlation analysis was conducted on the nine characteristic genes in relation to immune infiltration. The distribution of key genes in immune cells was determined using single-cell RNA sequencing data. Finally, the mRNA expression of the nine diagnostic genes was validated using qPCR.

The relationship between the extent of posterior limb of the internal capsule damage measured by non-contrast computed tomography and clinical outcomes after basal ganglia hemorrhage.

Assessing the extent of damage to the posterior limb of the internal capsule(PLIC) is important for early prediction of clinical outcomes in intracerebral hemorrhage (ICH) patients. Currently, using MRI to reconstruct the extent of damage to PLIC is not suitable for quick assessment of prognosis in emergency settings. We aimed to investigate whether the PLIC damage quantified by non-contrast computed tomography (NCCT) is associated with clinical outcomes after basal ganglia intracerebral hemorrhage (BG-ICH). This study retrospectively included 146 BG-ICH patients from the Department of Neurosurgery at the Second Affiliated Hospital of Chongqing Medical University. The damage to the PLIC was quantified using Tangency X measured by NCCT. The importance of features is determined using the Boruta algorithm and Least Absolute Shrinkage and Selection Operator (LASSO) regression. Multivariate logistic regression models were established to examine the impact of PLIC damage on outcomes. Restricted Cubic Splines (RCS) were used to explore potential nonlinear relationships, and Receiver Operating Characteristic (ROC) curves were used to compare the predictive performance of Tangency X with other scoring systems for 6-month neurological outcomes (poor outcomes [mRS: 3-6]). In the multivariate logistic regression adjusting for all covariates, Tangency X was independently associated with an increased risk of poor outcomes (OR = 1.32, 95% CI: 1.17-1.52) in BG-ICH patients. There is a nonlinear relationship between Tangency X and poor outcomes. Specifically, the risk of poor outcomes increases by 1.29 times (OR = 1.29, 95% CI: 1.09-1.67) for each additional 1mm increase in Tangency X beyond 4mm. We next observed that the AUC for Tangency X in predicting poor outcomes is 0.8511. The extent of PLIC damage measured by NCCT may represent a promising predictor of poor outcomes after BG-ICH.

Predictive Factors of Apparent Treatment Resistant Hypertension Among Patients With Hypertension Identified Using Electronic Health Records.

Early identification of a patient with resistant hypertension (RH) enables quickly intensified treatment, short-interval follow-up, or perhaps case management to bring his or her blood pressure under control and reduce the risk of complications. To identify predictors of RH among individuals with newly diagnosed hypertension (HTN), while comparing different prediction models and techniques for managing missing covariates using electronic health records data. Risk prediction study in a retrospective cohort. Adult patients with incident HTN treated in any of the primary care clinics of one health system between April 2013 and December 2016. Predicted risk of RH at the time of HTN identification and candidate predictors for variable selection in future model development. Among 26,953 individuals with incident HTN, 613 (2.3%) met criteria for RH after 4.7 months (interquartile range, 1.2-11.3). Variables selected by the least absolute shrinkage and selection operator (LASSO), included baseline systolic blood pressure (SBP) and its missing indicator (a dummy variable created if baseline SBP is absent), use of antihypertensive medication at the time of cohort entry, body mass index, and atherosclerosis risk. The random forest technique achieved the highest area under the curve (AUC) of 0.893 (95% CI, 0.881-0.904) and the best calibration with a calibration slope of 1.01. Complete case analysis is not a valuable option (AUC = 0.625). Machine learning techniques and traditional logistic regression exhibited comparable levels of predictive performance after handling the missingness. We suggest that the variables identified by this study may be good candidates for clinical prediction models to alert clinicians to the need for short-interval follow up and more intensive early therapy for HTN.

Integrating solar-induced chlorophyll fluorescence with traditional remote sensing and environmental variables for enhanced rice yield prediction in Nepal using machine learning

Rice is a vital cereal crop providing essential food for Nepal. To mitigate climate risks and ensure food security, especially with climate change and increasing extreme weather events, it is imperative and a prerequisite to have timely and reliable rice yield prediction. While traditional vegetation indices (VIs) and environmental variables are commonly used in rice yield prediction, the potential of solar-induced chlorophyll fluorescence (SIF), despite its greater sensitivity, has been explored in limited studies. Furthermore, integrating SIF with other remote sensing and environmental data for regional crop yield modeling still needs to be explored. This study aims to address these gaps by incorporating SIF to enhance rice yield prediction in Nepal, offering a novel approach to improving model accuracy. This study uses multi-source data and machine learning algorithms to study 22 major rice-growing districts in the Terai region of Nepal. Random forest (RF), support vector machine (SVR), gradient boosting regressor (GBR), and least absolute shrinkage and selection operator (LASSO) algorithms were employed to predict crop yield using various combinations of input features. Machine learning models were trained and tested over four different time windows within the growing period of rice, using 15 years of rice yield data from 2003 to 2017. Our analysis reveals that GBR out-performed other machine learning algorithms when using input features from July to September, achieving root mean square error (RMSE) and the ratio of RMSE (RRMSE) to observed mean values of 250 kg/ha and 8.6%, respectively. The findings also reveal that incorporating SIF alongside traditional remote sensing and environmental variables significantly enhances the model's prediction accuracy. These findings are valuable for ensuring food security, optimizing agricultural resource management, and supporting decision-making for farmers, policymakers, and supply chain stakeholders.

Machine learning and radiomics for ventricular tachyarrhythmia prediction in hypertrophic cardiomyopathy: insights from an MRI-based analysis.

Myocardial fibrosis is often detected in patients with hypertrophic cardiomyopathy (HCM), which causes left ventricular (LV) dysfunction and tachyarrhythmias. To evaluate the potential value of a machine learning (ML) approach that uses radiomic features from late gadolinium enhancement (LGE) and cine images for the prediction of ventricular tachyarrhythmia (VT) in patients with HCM. Hyperenhancing areas of LV myocardium on LGE images were manually segmented, and the segmentation was propagated to corresponding areas on cine images. Radiomic features were extracted using the PyRadiomics library. The least absolute shrinkage and selection operator (LASSO) method was employed for radiomic feature selection. Our model development employed the TabPFN algorithm, an adapted Prior-Data Fitted Network design. Model performance was evaluated graphically and numerically over five-repeat fivefold cross-validation. SHapley Additive exPlanations (SHAP) were employed to determine the relative importance of selected radiomic features. Our cohort consisted of 60 patients with HCM (73.3% male; median age = 51.5 years), among whom 17 had documented VT during the follow-up. A total of 1612 radiomic features were extracted for each patient. The LASSO algorithm led to a final selection of 18 radiomic features. The model achieved a mean area under the receiver operating characteristic curve of 0.877, demonstrating good discrimination, and a mean Brier score of 0.119, demonstrating good calibration. Radiomics-based ML models are promising for predicting VT in patients with HCM during the follow-up period. Developing predictive models as clinically useful decision-making tools may significantly improve risk assessment and prognosis.

Molecular subtypes and prognostic models for predicting prognosis of lung adenocarcinoma based on miRNA-related genes.

MicroRNAs (miRNAs) are crucial in cancer development and progression, and therapies targeting miRNAs demonstrate great therapeutic promise. We sought to predict the prognosis and therapeutic response of lung adenocarcinoma (LUAD) by classifying molecular subtypes and constructing a prognostic model based on miRNA-related genes. This study was based on miRNA-mRNA action pairs and ceRNA networks in the Cancer Genome Atlas (TCGA) database. Three molecular subtypes were determined based on 64 miRNA-associated target genes identified in the ceRNA network. The S3 subtype had the best prognosis, and the S2 subtype had the worst prognosis. The S2 subtype had a higher tumor mutational load (TMB) and a lower immune score. The S2 subtype was more suitable for immunotherapy and sensitive to chemotherapy. The least absolute shrinkage and selection operator (LASSO) algorithm was performed to determine eight miRNA-associated target genes for the construction of prognostic models. High-risk patients had a poorer prognosis, lower immune score, and lower response to immunotherapy. Robustness was confirmed in the Gene-Expression Omnibus (GEO) database cohort (GSE31210, GSE50081, and GSE37745 datasets). Overall, our study deepened the understanding of the mechanism of miRNA-related target genes in LUAD and provided new ideas for classification. Such miRNA-associated target gene characterization could be useful for prognostic prediction and contribute to therapeutic decision-making in LUAD.

Utility of multi-biomarker panel on discriminating disease activity in patients with psoriatic arthritis

ObjectiveTo investigate the correlation of serum protein biomarkers and disease activity in patients with PsA. Methods176 patients fulfilled the CASPAR (ClASsification criteria for Psoriatic ARthritis) were recruited in this cross-sectional study. The level of 48 protein biomarkers, cartilage and bone turn-over markers were assessed. The patients were randomly divided into a derivation-cohort and a validation-cohort at a ratio of 7:3. Patients were further categorized based on their disease activity states using cDAPSA (remission/low disease activity and moderate/high disease activity). Least absolute shrinkage and selection operator (LASSO) was used to select biomarkers which were associated with moderate/high disease activity in the derivation cohort. Receiver operating characteristic (ROC) curve, GiViTI calibration belt were used to assess the performance of the model in both cohorts. ResultsThe cohort [age: 55.5 (44.0–62.75) years, male: 80 (45.5 %)] had moderate disease activity [DAPSA: 15.9 (8.3–26.9); PASI: 3.2 (0.5–6.8)]. 101 PsA patients (57.4 %) had clinical DAPSA moderate/high disease activity. Biomarker levels associated with moderate/high disease activity included SAA (Serum amyloid A), IL-8 (Interleukin 8), IP10 (Interferon gamma-induced protein 10)/CXCL10, M-CSF (Macrophage colony-stimulating factor), SCGF-β (Stem cell growth factor), SDF-1α (Stromal cell-derived factor 1α)/CXCL12. The model’s equation including the 6 biomarker levels was applied to the validation-cohort. The area under the ROC curve (AUC) for discriminating moderate/high disease activity was 0.802 and 0.835 for the derivation-and-validation-cohorts, respectively. The multi-biomarkers panel model had higher-AUC when compared with that of C-reactive protein (CRP) (AUC = 0.727, p = 0.022). The P-values of calibration charts in the two sets were 0.902 and 0.123. ConclusionsThe multi-biomarkers panel demonstrated the ability to discriminate patients with moderate/high disease activity from those with low disease activity/remission.

Identification of diagnostic candidate genes in COVID-19 patients with sepsis.

Coronavirus Disease 2019 (COVID-19) and sepsis are closely related. This study aims to identify pivotal diagnostic candidate genes in COVID-19 patients with sepsis. We obtained a COVID-19 data set and a sepsis data set from the Gene Expression Omnibus (GEO) database. Identification of differentially expressed genes (DEGs) and module genes using the Linear Models for Microarray Data (LIMMA) and weighted gene co-expression network analysis (WGCNA), functional enrichment analysis, protein-protein interaction (PPI) network construction, and machine learning algorithms (least absolute shrinkage and selection operator (LASSO) regression and Random Forest (RF)) were used to identify candidate hub genes for the diagnosis of COVID-19 patients with sepsis. Receiver operating characteristic (ROC) curves were developed to assess the diagnostic value. Finally, the data set GSE28750 was used to verify the core genes and analyze the immune infiltration. The COVID-19 data set contained 3,438 DEGs， and 595 common genes were screened in sepsis. sepsis DEGs were mainly enriched in immune regulation. The intersection of DEGs for COVID-19 and core genes for sepsis was 329, which were also mainly enriched in the immune system. After developing the PPI network, 17 node genes were filtered and thirteen candidate hub genes were selected for diagnostic value evaluation using machine learning. All thirteen candidate hub genes have diagnostic value, and 8 genes with an Area Under the Curve (AUC) greater than 0.9 were selected as diagnostic genes. Five core genes (CD3D, IL2RB, KLRC, CD5, and HLA-DQA1) associated with immune infiltration were identified to evaluate their diagnostic utility COVID-19 patients with sepsis. This finding contributes to the identification of potential peripheral blood diagnostic candidate genes for COVID-19 patients with sepsis.

Integrating intratumoral and peritumoral radiomics with clinical risk factors for prognostic prediction in pancreatic ductal adenocarcinoma patients undergoing combined chemotherapy and HIFU ablation

Objective A radiomics nomogram will be created utilizing MRI data from intratumoral and peritumoral areas to forecast survival outcomes in patients who have had treatment for pancreatic ductal adenocarcinoma (PDAC). Methods A total of 87 individuals diagnosed with PDAC were included in the study, with 60 patients in the training cohort and 27 patients in the validation cohort. A grand total of 2395 radiomics characteristics were extracted from the tumor region and the peritumoral region. The least absolute shrinkage and selection operator (LASSO) method was used to select features and create a radiomics score, also known as the Rad-score. A multivariate regression analysis was then conducted to build the radiomics nomogram. The evaluation of the nomogram included discrimination, calibration, and clinical utility assessments. Results Based on the conclusions derived from the multivariate Cox model, Rad-Score, jaundice, and tumor size were identified as independent risk factors for overall survival (OS). The inclusion of the Rad-score in the radiomics nomogram led to improved accuracy in predicting survival compared to the clinical model. Patients were categorized into high-risk and low-risk groups based on their Rad-Score. Kaplan-Meier analysis revealed a statistically significant difference between the two groups (p < 0.05). Furthermore, the radiomics nomogram demonstrated excellent ability to differentiate, calibrate, and provide clinical utility in both the training and validation groups. Conclusions The MRI-based intratumoral and peritumoral radiomics nomogram, integrating the Rad-score and clinical data, provided better prognostic prediction for PDAC patients after HIFU treatment, which may hold great potential for guiding personalized care for these patients.

Decoding the immune microenvironment: unveiling CD8 + T cell-related biomarkers and developing a prognostic signature for personalized glioma treatment

BackgroundGliomas are aggressive brain tumors with poor prognosis. Understanding the tumor immune microenvironment (TIME) in gliomas is essential for developing effective immunotherapies. This study aimed to identify TIME-related biomarkers in glioma using bioinformatic analysis of RNA-seq data.MethodsIn this study, we employed weighted gene co-expression network analysis (WGCNA) on bulk RNA-seq data to identify TIME-related genes. To identify prognostic genes, we performed univariate Cox regression and least absolute shrinkage and selection operator (LASSO) regression analyses. Based on these genes, we constructed a prognostic signature and delineated risk groups. To validate the prognostic signature, external validation was conducted.ResultsCD8 + T cell infiltration was strongly correlated with glioma patient prognosis. We identified 115 CD8 + T cell-related genes through integrative analysis of bulk-seq data. CDCA5, KIF11, and KIF4A were found to be significant immune-related genes (IRGs) associated with overall survival in glioma patients and served as independent prognostic factors. We developed a prognostic nomogram that incorporated these genes, age, gender, and grade, providing a reliable tool for clinicians to predict patient survival probabilities. The nomogram’s predictions were supported by calibration plots, further validating its accuracy.ConclusionIn conclusion, our study identifies CD8 + T cell infiltration as a strong predictor of glioma patient outcomes and highlights the prognostic value of genes. The developed prognostic nomogram, incorporating these genes along with clinical factors, provides a reliable tool for predicting patient survival probabilities and has important implications for personalized treatment decisions in glioma.

Associations between exposure to multiple environmental chemicals and metabolic syndrome: A mixture analysis

Exposure to environmental chemicals is prevalent. While previous studies reported associations between multiple chemical exposures and metabolic syndrome (MetS), they did not comprehensively account for correlations among exposures. We used machine learning methods including Boruta algorithm and least absolute shrinkage and selection operator (LASSO) regression, combined with weighted quartiles sum (WQS) regression to investigate the associations of phenols, polycyclic aromatic hydrocarbons (PAHs), metals, and phthalates with MetS and its components. Data were drawn from the 2005–2012 National Health and Nutrition Examination Survey (NHANES). The mean (standard deviation (SD)) age of 2596 participants was 48.4 (17.9) years. After adjusting for age, sex, body mass index, race/ethnicity, marital status, education, poverty income ratio, physical activity, smoking, and alcohol, higher 2-Phenanthrene (2-PHE) and mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP) concentrations were associated with a higher odds of the MetS (odds ratio (OR) = 4.26, 95 % confidence interval (CI) 2.40–7.58 per ng/mL, and 3.24, 1.75–6.02 per ng/L, respectively). WQS index for environmental chemicals was positively associated with the MetS (OR = 1.31, 95 % CI 1.09–1.57). Moreover, we observed consistent and stronger positive associations with MetS (OR = 1.54, 95 % CI 1.04–2.30) in current smokers. Exposure to phenols, PAHs, metals, and phthalates was positively associated with an increase in metabolic syndrome and its components, which was more pronounced in current smokers.

Identification and RT-qPCR Validation of Biomarkers Based on Butyrate Metabolism-Related Genes to Predict Recurrent Miscarriage.

To date, the cause of recurrent miscarriage (RM) in at least 50% of patients remains unknown. However, no study has explored the correlation between butyrate metabolism-related genes (BMRGs) and RM. RM-related datasets (GSE165004, GSE111974, GSE73025, and GSE179996) were obtained from the Gene Expression Omnibus (GEO) database. First, 595 differentially expressed genes (DEGs) were identified between the RM and control samples in GSE165004. Subsequently, 213 differentially expressed BMRGs (DE-BMRGs) were identified by considering the intersection of DEGs with BMRGs. The protein-protein interaction (PPI)network of DE-BMRGs contained 156 nodes and 250 edges, and a key module was obtained. In total, four biomarkers (ACTR2, ANXA2, PFN1, and OAS1) were acquired through least absolute shrinkage and selection operator (LASSO), support vector machine-recursive feature elimination (SVM-RFE), and random forest (RF). Immune analysis revealed two immune cells and three immune-related gene sets that were significantly different between the RM and control groups, namely, T helper cells, regulatory T cells (Treg), MHC class I, parainflammation, and type I IFN response. In addition, a TF-mRNA network based on the top 100 nodes ranked in the order of connectivity was created, including 100 nodes and 253 edges, such as MTERF2-ACTR2, NKX23-PFN1, STAT1-OAS1, and SP100-ANXA2. Finally, 3 drugs (withaferin A, N-ethylmaleimide, and etoposide) were predicted to interact with 2 biomarkers (ANXA2 and ACTR2). Eventually, ANXA2 and OAS1 were significantly downregulated, and PFN1 was markedly overexpressed in the RM group, as determined by reverse transcription quantitative polymerase chain reaction (RT-qPCR). Our findings authenticated four butyrate metabolism-related biomarkers for the diagnosis of RM, providing a scientific reference for further studies on RM treatment.

Radiomics model for predicting distant metastasis in soft tissue sarcoma of the extremities and trunk treated with surgery.

The aim of this study was to develop a radiomics model based on magnetic resonance imaging (MRI) for predicting metastasis in soft tissue sarcomas (STSs) treated with surgery. MRI and clinical data of 73 patients with STSs of the extremities and trunk were obtained from TCIA database and Jiangsu Cancer Hospital as the training set, data of other 40 patients were retrospectively collected at our institution as the external validation set. Radiomics features were extracted from both intratumoral and peritumoral regions of fat-suppressed T2-weighted images (FS-T2WIs) of patients, and 3D ResNet10 was used to extract deep learning features. Recursive feature elimination (RFE) and least absolute shrinkage and selection operator (LASSO) algorithms were used for the selection of features. Based on 4 different sets of features, 5 machine learning algorithms were used to construct intratumor, peritumor, combined intratumor and peritumor radiomics models and deep learning radiomics (DLR) model. The area under the ROC curve (AUC) and Decision curve analysis (DCA) were used to evaluate the ability of models to predict metastasis. Based on 20 selected features from the deep-learning and radiomics features set, the DLR model was able to predict metastasis in the validation dataset, with an AUC of 0.9770. The DCA and Hosmer-Lemeshow test revealed that the DLR model had good clinical benefit and consistency. By getting richer information from MRI, The DLR model is a noninvasive, low-cost method for predicting the risk of metastasis in STSs, and can help develop appropriate treatment programs.

Prediction and validation of pathologic complete response for locally advanced rectal cancer under neoadjuvant chemoradiotherapy based on a novel predictor using interpretable machine learning

BackgroundPrecise evaluation of pathological complete response (pCR) is essential for determining the prognosis of patients with locally advanced rectal cancer (LARC) undergoing neoadjuvant chemoradiotherapy (NCRT) and can offer clues for the selection of subsequent treatment strategies. Most current predictive models for pCR focus primarily on pre-treatment factors, neglecting the dynamic systemic changes that occur during neoadjuvant chemoradiotherapy, and are constrained by low accuracy and lack of integrity. PurposeThis study devised a novel predictor of pCR using dynamic alterations in systemic inflammation-nutritional marker indexes (SINI) during neoadjuvant therapy and developed a machine-learning model to predict pCR. MethodsTwo cohorts of patients with LARC from center one from 2012 to 2017 and from center two from 2020 to 2023 were integrated for analysis. This study compared dynamic changes in blood indexes before and after neoadjuvant therapy and surgical operation. A least absolute shrinkage and selection operator (LASSO) regression analysis was conducted to mitigate collinearity and identify key indexes, constructing the SINI. Univariate and multiple logistic regression analyses were used to identify the independent risk factors associated with pCR. Additionally, 10 machine learning algorithms were employed to develop predictive models to assess risk. The hyperparameters of the machine learning models were optimized using a random search and 10-fold cross-validation. The models were assessed by examining various metrics, including the area under the receiver operating characteristic curves (AUC), the area under the precision-recall curve (AUPRC), decision curve analysis, calibration curves, and the precision and accuracy of the internal and external validation cohorts. Additionally, Shapley’s additive explanations (SHAP) were employed to interpret the machine learning models. ResultsThe study cohort comprised 677 patients from the center one and 224 patients from the center two. Six key indexes were identified, and a predictive index, SINI, was constructed. Univariate and multiple logistic regression analyses revealed that SINI, clinical T-stage, clinical N-stage, tumor size, and the distance from the anal verge were independent risk factors for pCR in patients with LARC following NCRT. The mean AUC value of the extreme gradient boosting (XGB) model in the 10-fold cross-validation of the training set was 0.877. The XGB model demonstrated superior performance in the internal and external validation sets. Specifically, in the internal test set, the XGB model achieved an AUC of 0.86, AUPRC of 0.707, accuracy of 0.82, and precision of 0.80. In the external validation set, the XGB model exhibited an AUC of 0.83, AUPRC of 0.702, accuracy of 0.81, and precision of 0.81. Additionally, the predictions generated by the XGB model were analyzed using SHAP. ConclusionThis study involved developing and validating an XGB model using SINI to predict pCR in patients with LARC. Besides, a SINI-based machine learning model shows promise in accurately predicting pCR following NCRT in patients with resectable LARC, offering valuable insights for personalized treatment approaches.

Integrating Bioinformatics and Experimental Validation to Identify Mitochondrial Permeability Transition-Driven Necrosis-Related lncRNAs that can Serve as Prognostic Biomarkers and Therapeutic Targets in Endometrial Carcinoma.

Endometrial carcinoma (EC) is a common malignant tumor in women with high mortality and relapse rates. Mitochondrial permeability transition (MPT)-driven necrosis is a novel form of programmed cell death. The MPT-driven necrosis related lncRNAs (MRLs) involved in EC development remain unclear. We aimed to predict the outcomes of patients with EC by constructing a novel prognostic model based on MRLs and explore potential molecular functions. A risk prognostic model was developed utilizing multi-Cox regression in conjunction with the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithm, which was based on MRLs. The predictive efficacy of the model was evaluated through receiver operating characteristic (ROC) curve analysis, as well as nomogram and concordance index (C-index) assessments. Patients were categorized into high- and low-risk groups based on their median risk scores. Notably, the high-risk group exhibited significantly poorer overall survival (OS) outcomes. Gene ontology (GO) and Gene set enrichment analysis (GSEA) demonstrated that Hedgehog and cell cycle pathways were enriched in the high-risk group. Tumor Immune Dysfunction and Exclusion (TIDE) displayed that patients in the high-risk group showed a high likelihood of immune evasion and less effective immunotherapy. A significant disparity in immune function was also observed between two groups. Based on the nine-MRLs, drug sensitivity analysis identified several anticancer drugs with potential efficacy in prognosis. Meanwhile, the results demonstrated that OGFRP1 plays a carcinogenic role by affecting mitochondrial membrane permeability in EC. Therefore, the risk model constructed by nine MRLs could be used to predict the clinical outcomes and therapeutic responses in patients with EC effectively.

Bayesian estimation and comparison of idiographic network models.

Idiographic network models are estimated on time series data of a single individual and allow researchers to investigate person-specific associations between multiple variables over time. The most common approach for fitting graphical vector autoregressive (GVAR) models uses least absolute shrinkage and selection operator (LASSO) regularization to estimate a contemporaneous and a temporal network. However, estimation of idiographic networks can be unstable in relatively small data sets typical for psychological research. This bears the risk of misinterpreting differences in estimated networks as spurious heterogeneity between individuals. As a remedy, we evaluate the performance of a Bayesian alternative for fitting GVAR models that allows for regularization of parameters while accounting for estimation uncertainty. We also develop a novel test, implemented in the tsnet package in R, which assesses whether differences between estimated networks are reliable based on matrix norms. We first compare Bayesian and LASSO approaches across a range of conditions in a simulation study. Overall, LASSO estimation performs well, while a Bayesian GVAR without edge selection may perform better when the true network is dense. In an additional simulation study, the novel test is conservative and shows good false-positive rates. Finally, we apply Bayesian estimation and testing in an empirical example using daily data on clinical symptoms for 40 individuals. We additionally provide functionality to estimate Bayesian GVAR models in Stan within tsnet. Overall, Bayesian GVAR modeling facilitates the assessment of estimation uncertainty which is important for studying interindividual differences of intraindividual dynamics. In doing so, the novel test serves as a safeguard against premature conclusions of heterogeneity. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Identifying Brain Network Structure for an fMRI Effective Connectivity Study Using the Least Absolute Shrinkage and Selection Operator (LASSO) Method

Background: Studying causality relationships between different brain regions using the fMRI method has attracted great attention. To investigate causality relationships between different brain regions, we need to identify both the brain network structure and the influence magnitude. Most current methods concentrate on magnitude estimation, but not on identifying the connection or structure of the network. To address this problem, we proposed a nonlinear system identification method, in which a polynomial kernel was adopted to approximate the relation between the system inputs and outputs. However, this method has an overfitting problem for modelling the input–output relation if we apply the method to model the brain network directly. Methods: To overcome this limitation, this study applied the least absolute shrinkage and selection operator (LASSO) model selection method to identify both brain region networks and the connection strength (system coefficients). From these coefficients, the causality influence is derived from the identified structure. The method was verified based on the human visual cortex with phase-encoded designs. The functional data were pre-processed with motion correction. The visual cortex brain regions were defined based on a retinotopic mapping method. An eight-connection visual system network was adopted to validate the method. The proposed method was able to identify both the connected visual networks and associated coefficients from the LASSO model selection. Results: The result showed that this method can be applied to identify both network structures and associated causalities between different brain regions. Conclusions: System identification with LASSO model selection algorithm is a powerful approach for fMRI effective connectivity study.

A Robust [18F]-PSMA-1007 Radiomics Ensemble Model for Prostate Cancer Risk Stratification.

The aim of this study is to investigate the role of [18F]-PSMA-1007 PET in differentiating high- and low-risk prostate cancer (PCa) through a robust radiomics ensemble model. This retrospective study included 143 PCa patients who underwent [18F]-PSMA-1007 PET/CT imaging. PCa areas were manually contoured on PET images and 1781 image biomarker standardization initiative (IBSI)-compliant radiomics features were extracted. A 30 times iterated preliminary analysis pipeline, comprising of the least absolute shrinkage and selection operator (LASSO) for feature selection and fivefold cross-validation for model optimization, was adopted to identify the most robust features to dataset variations, select candidate models for ensemble modelling, and optimize hyperparameters. Thirteen subsets of selected features, 11 generated from the preliminary analysis plus two additional subsets, the first based on the combination of robust and fine-tuning features, and the second only on fine-tuning features were used to train the model ensemble. Accuracy, area under curve (AUC), sensitivity, specificity, precision, and f-score values were calculated to provide models' performance. Friedman test, followed by post hoc tests corrected with Dunn-Sidak correction for multiple comparisons, was used to verify if statistically significant differences were found in the different ensemble models over the 30 iterations. The model ensemble trained with the combination of robust and fine-tuning features obtained the highest average accuracy (79.52%), AUC (85.75%), specificity (84.29%), precision (82.85%), and f-score (78.26%). Statistically significant differences (p < 0.05) were found for some performance metrics. These findings support the role of [18F]-PSMA-1007 PET radiomics in improving risk stratification for PCa, by reducing dependence on biopsies.