EPS-LASSO: test for high-dimensional regression under extreme phenotype sampling of continuous traits

Metaheuristic integrated machine learning classification of colon cancer using STFT LASSO and EHO feature extraction from microarray gene expressions

The microarray gene expression data poses a tremendous challenge due to their curse of dimensionality problem. The sheer volume of features far surpasses available samples, leading to overfitting and reduced classification accuracy. Thus the dimensionality of microarray gene expression data must be reduced with efficient feature extraction methods to reduce the volume of data and extract meaningful information to enhance the classification accuracy and interpretability. In this research, we discover the uniqueness of applying STFT (Short Term Fourier Transform), LASSO (Least Absolute Shrinkage and Selection Operator), and EHO (Elephant Herding Optimisation) for extracting significant features from lung cancer and reducing the dimensionality of the microarray gene expression database. The classification of lung cancer is performed using the following classifiers: Gaussian Mixture Model (GMM), Particle Swarm Optimization (PSO) with GMM, Detrended Fluctuation Analysis (DFA), Naive Bayes classifier (NBC), Firefly with GMM, Support Vector Machine with Radial Basis Kernel (SVM-RBF) and Flower Pollination Optimization (FPO) with GMM. The EHO feature extraction with the FPO-GMM classifier attained the highest accuracy in the range of 96.77, with an F1 score of 97.5, MCC of 0.92 and Kappa of 0.92. The reported results underline the significance of utilizing STFT, LASSO, and EHO for feature extraction in reducing the dimensionality of microarray gene expression data. These methodologies also help in improved and early diagnosis of lung cancer with enhanced classification accuracy and interpretability.

Immune-related eight-lncRNA signature for improving prognosis prediction of lung adenocarcinoma

Background

Lung adenocarcinoma (LUAD) is the leading cause of cancer‐related deaths worldwide. Therefore, the identification of a novel prediction signature for predicting the prognosis risk and survival outcomes is urgently demanded.

Methods

We integrated a machine‐learning frame by combing the Cox regression and Least Absolute Shrinkage and Selection Operator (LASSO) regression model to identify the LUAD‐related long non‐coding RNA (lncRNA) survival biomarkers. Subsequently, the Spearman correlation test was employed to interrogate the relationships between lncRNA signature and tumor immunity and constructed the competing endogenous RNA (ceRNA) network.

Results

Herein, we identified an eight‐lncRNA signature (PR‐lncRNA signature, NPSR1‐AS1, SATB2‐AS1, LINC01090, FGF12‐AS2, AC005256.1, MAFA‐AS1, BFSP2‐AS1, and CPC5‐AS1), which contributes to predicting LUAD patient's prognosis risk and survival outcomes. The PR‐lncRNA signature has also been confirmed as the robust signature in independent datasets. Further parsing of the LUAD tumor immune infiltration showed the PR‐lncRNAs were closely associated with the abundance of multiple immune cells infiltration and the expression of MHC molecules. Furthermore, by constructing the PR‐lncRNA–related ceRNA network, we interrogated more potential anti‐cancer therapy targets.

Conclusion

lncRNAs, as emerging cancer biomarkers, play an important role in a variety of cancer processes. Identification of PR‐lncRNA signatures allows us to better predict patient's survival outcomes and disease risk. Finally, the PR‐lncRNA signatures could help us to develop novel LUAD anti‐cancer therapeutic strategies.

A Panel of Five-lncRNA Signature as a Potential Biomarker for Predicting Survival in Gastric and Thoracic Cancers

Dysfunctional long non-coding RNAs (lncRNAs) have been found to have carcinogenic and/or tumor inhibitory effects in the development and progression of cancer, suggesting their potential as new independent biomarkers for cancer diagnosis and prognosis. The exploration of the relationship between lncRNAs and the overall survival (OS) of different cancers opens up new prospects for tumor diagnosis and treatment. In this study, we established a five-lncRNA signature and explored its prognostic efficiency in gastric cancer (GC) and several thoracic malignancies, including breast invasive carcinoma (BRCA), esophageal carcinoma, lung adenocarcinoma, lung squamous cell carcinoma (LUSC), and thymoma (THYM). Cox regression analysis and lasso regression were used to evaluate the relationship between lncRNA expression and survival in different cancer datasets from GEO and TCGA. Kaplan-Meier survival curves indicated that risk scores characterized by a five-lncRNA signature were significantly associated with the OS of GC, BRCA, LUSC, and THYM patients. Functional enrichment analysis showed that these five lncRNAs are involved in known biological pathways related to cancer pathology. In conclusion, the five-lncRNA signature can be used as a prognostic marker to promote the diagnosis and treatment of GC and thymic malignancies.

Novel physiologic nomogram discriminates symptom outcome in patients with erosive esophagitis

BACKGROUND AND AIM

Most of patients with erosive esophagitis (EE) are of LA grade A&B with low reflux burden, therefore require further esophageal function tests (EFTs). One-third of them respond poorly to pump proton inhibitor (PPI) treatment. The aim was to establish and validate a physiologic nomogram to discriminate symptom outcome to PPI treatment in patients with EE.

METHODS

A total of 79 EE patients with heartburn who underwent EFTs and received PPI therapy were randomly assigned into a training set (n = 55) and a validation set (n = 24). Clinical data including physiologic parameters from EFTs were collected. Significant factors for the positive symptomatic outcome were identified using logistic regression analysis. Physiologic signature was developed using the least absolute shrinkage and selection operator algorithm. The nomogram was established by combining significant factors and physiologic signature, and its performance was evaluated and validated in the training and validation set. The clinical value of the nomogram was measured by decision curve analysis.

RESULTS

Significant factors for positive symptomatic response to PPI treatment were identified as follows: acid exposure time, total number of reflux episodes, and two novel metrics including mean nocturnal baseline impedance and post-reflux swallow-induced peristaltic wave index. The nomogram which incorporated both significant factors and physiologic signature demonstrated good performance in the training and validation sets [C-index: 0.938 (95% CI 0.882-0.995); 0.839 (95% CI 0.678-0.995), respectively]. Decision curves showed significant clinical usefulness.

CONCLUSION

The first physiologic nomogram was developed to discriminate the individualized response to PPI therapy among EE patients.

Extreme phenotypes approach to investigate host genetics and COVID-19 outcomes

COVID-19 comprises clinical outcomes of SARS-CoV-2 infection and is highly heterogeneous, ranging from asymptomatic individuals to deceased young adults without comorbidities. There is growing evidence that host genetics play an important role in COVID-19 severity, including inborn errors of immunity, age-related inflammation and immunosenescence. Here we present a brief review on the known order of events from infection to severe system-wide disturbance due to COVID-19 and summarize potential candidate genes and pathways. Finally, we propose a strategy of subject's ascertainment based on phenotypic extremes to take part in genomic studies and elucidate intrinsic risk factors involved in COVID-19 severe outcomes.

Transition From Acute to Chronic Pain in Lower Extremity Fracture Patients: A Pain Phenotyping Protocol

BACKGROUND

Traumatic injury is a major source of chronic pain, particularly for individuals with traumatic fracture of the fibula and/or tibia (lower extremity fracture [LEFx]). Although several factors (e.g., older age, being female, high pain intensity at time of initial injury) have been identified as risk factors for chronic pain associated with LEFx. Comprehensive biopsychosical models to predict the odds of transitioning from acute to chronic pain after LEFx are needed to better understand the underlying processes, predict risk for chronic pain, and develop personalized therapies for individuals at higher risk for developing chronic pain.

OBJECTIVE

The aim of the study was to outline the study design that will be used to examine the physiological, psychological, and genetic/genomic variables-models that predict the transition from acute to chronic pain after LEFx.

METHOD

This prospective descriptive cohort study will enroll 240 participants with a fibula and/or tibia fracture and 40 controls with no LEFx. Data will be collected during an in-hospital baseline visit, five in-person clinic visits (6 weeks, 12 weeks, 24 weeks, 52 weeks, and 24 months), and seven online between-visit surveys (2 weeks, 4 weeks, 8 weeks, 10 weeks, 16 weeks, 20 weeks, and 18 months) from participants with LEFx and at concordant intervals from controls. Measures will consist of 19 questionnaires characterizing pain and psychological status, neurophysiological testing for peripheral sensory nerve function, and peripheral blood samples collections for RNA sequencing. Illumina standard protocols will be used to sequence RNA, and read counts will be used to measure gene expression.

ANALYSIS

Direct-entry, multiple logistic regression will be used to produce odds ratios expressing the relative risk on each explanatory variable when controlling for other predictors/covariates in the model.

CONCLUSION

This study is one of the first to longitudinally characterize the biopsychosocial variables associated with a clinically relevant problem of the transition from acute to chronic posttraumatic fracture pain in individuals with LEFx. Results from this study will be used to construct predictive risk models of physiological, psychological, and genetic/genomic variables associated with increased risk for transitioning from acute to chronic pain status after LEFx. This work will lead to a better understanding of the trajectory of pain and relevant variables over time; initiate a better understanding of variables associated with risk for transitioning from acute to chronic pain; and, in the future, could provide a foundation for the identification of novel therapeutic targets to improve the outcomes of individuals with LEFx.

Integrated Nomograms for Preoperative Prediction of Microvascular Invasion and Lymph Node Metastasis Risk in Hepatocellular Carcinoma Patients

BACKGROUND

The aim of the present work is to develop and validate accurate preoperative nomograms to predict microvascular invasion (MVI) and lymph node metastasis (LNM) in hepatocellular carcinoma.

PATIENTS AND METHODS

A total of 268 patients with resected hepatocellular carcinoma (HCC) were divided into a training set (n = 180), in an earlier period, and a validation set (n = 88), thereafter. Risk factors for MVI and LNM were assessed based on logistic regression. Blood signatures were established using the least absolute shrinkage and selection operator algorithm. Nomograms were constructed by combining risk factors and blood signatures. Performance was evaluated using the training set and validated using the validation set. The clinical values of the nomograms were measured by decision curve analysis.

RESULTS

The risk factors for MVI were hepatitis B virus (HBV) DNA loading, portal hypertension, Barcelona liver clinic (BCLC) stage, and three computerized tomography (CT) imaging features, namely tumor number, size, and encapsulation, while only BCLC stage, Child-Pugh classification, and tumor encapsulation were associated with LNM. The nomogram incorporating both risk factors and blood signatures achieved better performance in predicting MVI in the training and validation sets (C-indexes of 0.828 and 0.804) than the LNM nomogram (C-indexes of 0.765 and 0.717). Calibration curves also demonstrated a good fit. The decision curves indicate significant clinical usefulness.

CONCLUSIONS

The novel validated nomograms for HCC patients presented herein are noninvasive preoperative tools that can effectively predict the individualized risk of MVI and LNM, and this predictive power can aid doctors in explaining the illness for patient counseling.

Identification and validation of a prognostic four-genes signature for hepatocellular carcinoma: integrated ceRNA network analysis

BACKGROUND

Hepatocellular carcinoma (HCC) is one of the most aggressive malignant tumors, with a poor long-term prognosis worldwide. The functional deregulations of global transcriptome were associated with the genesis and development of HCC, but lacks systematic research and validation.

METHODS

A total of 519 postoperative HCC patients were included. We built an interactive and visual competing endogenous RNA network. The prognostic signature was established with the least absolute shrinkage and selection operator algorithm. Multivariate Cox regression analysis was used to screen for independent prognostic factors for HCC overall survival.

RESULTS

In the training set, we identified a four-gene signature (PBK, CBX2, CLSPN, and CPEB3) and effectively predicted the overall survival. The survival times of patients in the high-score group were worse than those in the low-score group (p = 0.0004), and death was also more likely in the high-score group (HR 2.444, p < 0.001). The results were validated in internal validation set (p = 0.0057) and two external validation cohorts (HR 2.467 and 2.6). The signature (AUCs of 1, 2, 3 years were 0.716, 0.726, 0.714, respectively) showed high prognostic accuracy in the complete TCGA cohort.

CONCLUSIONS

In conclusion, we successfully built a more extensive ceRNA network for HCC and then identified a four-gene-based signature, enabling prediction of the overall survival of patients with HCC.

An accurate regression of developmental stages for breast cancer based on transcriptomic biomarkers

AIM

Breast cancers at different stages have tremendous differences on both phenotypic and molecular patterns. The developmental stage is an essential factor in the clinical decision of treatment plans, but was usually formulated as a classification problem, which ignored the consecutive relationships among them.

MATERIALS & METHODS

This study proposed a regression-based procedure to detect the stage biomarkers of breast cancers. Biomarkers were detected by the Lasso and Ridge algorithms.

RESULTS & CONCLUSION

A collaboration duet of Lasso and Ridge regression algorithms achieved the best performances, with classification accuracy (Acc) equal to 0.8294 and regression goodness-of-fit (R²) equal to 0.7810. The 265 biomarker genes were enriched with the signal peptide-based secretion function with the Bonferroni-corrected p-value equal to 6.9408e-3 and false discovery rate (FDR) equal to 1.1614e-2.

Association of Plasma Irisin with Bone Mineral Density in a Large Chinese Population Using an Extreme Sampling Design

Irisin, a myokine produced by skeletal muscle in response to physical exercise, promotes trans-differentiation of white adipose tissue into brown adipose tissue. Recent evidences suggested that irisin also plays an important role in the control of bone metabolism. This study aimed to ascertain the relationship between plasma irisin and bone mineral density (BMD) in Chinese population by adoption of an extreme sampling method. Based on a large and screened Chinese elderly population (N = 6308), two subgroups with extremely high and low hip BMD were selected for discovery (N = 80, high vs. low BMD = 44:36) and validation (N = 60, high vs. low BMD = 30:30), respectively. Plasma irisin, P1NP, and β-CTx were measured using commercially available ELISA kits. Other metabolic parameters (e.g., blood glucose, total cholesterol and triglycerides) were collected. Student's t test and Spearman correlation analyses were conducted in SPSS. Significant difference was discovered for plasma irisin between females and age-matched males (N = 80, male vs. female = 42:38, P = 0.002). The plasma irisin levels were significantly higher in high BMD subjects than in low BMD subjects, which was observed in both discovery (P = 0.012) and validation samples (P = 0.022). However, such observation was limited to males only. Further correlation analyses in males showed that plasma irisin was correlated with BMD (r = 0.362, P = 0.025) and triglyceride (r = - 0.354, P = 0.032). Plasma irisin levels were associated with hip BMD in Chinese elderly men. This study represented the first effort of investigating the relationship of plasma irisin and BMD in elderly population. The positive correlation between plasma irisin and BMD hints intrinsic communication between muscle and bone.