Optimizing cancer classification: a hybrid RDO-XGBoost approach for feature selection and predictive insights.

The identification of relevant biomarkers from high-dimensional cancer data remains a significant challenge due to the complexity and heterogeneity inherent in various cancer types. Conventional feature selection methods often struggle to effectively navigate the vast solution space while maintaining high predictive accuracy. In response to these challenges, we introduce a novel feature selection approach that integrates Random Drift Optimization (RDO) with XGBoost, specifically designed to enhance the performance of cancer classification tasks. Our proposed framework not only improves classification accuracy but also offers valuable insights into the underlying biological mechanisms driving cancer progression. Through comprehensive experiments conducted on real-world cancer datasets, including Central Nervous System (CNS), Leukemia, Breast, and Ovarian cancers, we demonstrate the efficacy of our method in identifying a smaller subset of unique and relevant genes. This selection results in significantly improved classification efficiency and accuracy. When compared with popular classifiers such as Support Vector Machine, K-Nearest Neighbor, and Naive Bayes, our approach consistently outperforms these models in terms of both accuracy and F-measure metrics. For instance, our framework achieved an accuracy of 97.24% in the CNS dataset, 99.14% in Leukemia, 95.21% in Ovarian, and 87.62% in Breast cancer, showcasing its robustness and effectiveness across different types of cancer data. These results underline the potential of our RDO-XGBoost framework as a promising solution for feature selection in cancer data analysis, offering enhanced predictive performance and valuable biological insights.

Insights into the Gene Expression Profile of Classical Hodgkin Lymphoma: A Study towards Discovery of Novel Therapeutic Targets.

Classical Hodgkin lymphoma (cHL) is a common B-cell cancer and a significant health concern, especially in Western and Asian countries. Despite the effectiveness of chemotherapy, many relapse cases are being reported, highlighting the need for improved treatments. This study aimed to address this issue by discovering biomarkers through the analysis of gene expression data specific to cHL. Additionally, potential anticancer inhibitors were explored to target the discovered biomarkers. This study proceeded by retrieving microarray gene expression data from cHL patients, which was then analyzed to identify significant differentially expressed genes (DEGs). Functional and network annotation of the upregulated genes revealed the active involvement of matrix metallopeptidase 12 (MMP12) and C-C motif metallopeptidase ligand 22 (CCL22) genes in the progression of cHL. Additionally, the mentioned genes were found to be actively involved in cancer-related pathways, i.e., oxidative phosphorylation, complement pathway, myc_targets_v1 pathway, TNFA signaling via NFKB, etc., and showed strong associations with other genes known to promote cancer progression. MMP12, topping the list with a logFC value of +6.6378, was selected for inhibition using docking and simulation strategies. The known anticancer compounds were docked into the active site of the MMP12 molecular structure, revealing significant binding scores of -7.7 kcal/mol and -7.6 kcal/mol for BDC_24037121 and BDC_27854277, respectively. Simulation studies of the docked complexes further supported the effective binding of the ligands, yielding MMGBSA and MMPBSA scores of -78.08 kcal/mol and -82.05 kcal/mol for MMP12-BDC_24037121 and -48.79 kcal/mol and -49.67 kcal/mol for MMP12-BDC_27854277, respectively. Our findings highlight the active role of MMP12 in the progression of cHL, with known compounds effectively inhibiting its function and potentially halting the advancement of cHL. Further exploration of downregulated genes is warranted, as associated genes may play a role in cHL. Additionally, CCL22 should be considered for further investigation due to its significant role in the progression of cHL.

Genetic algorithm-based feature selection with manifold learning for cancer classification using microarray data.

BACKGROUND: Microarray data have been widely utilized for cancer classification. The main characteristic of microarray data is "large p and small n" in that data contain a small number of subjects but a large number of genes. It may affect the validity of the classification. Thus, there is a pressing demand of techniques able to select genes relevant to cancer classification. RESULTS: This study proposed a novel feature (gene) selection method, Iso-GA, for cancer classification. Iso-GA hybrids the manifold learning algorithm, Isomap, in the genetic algorithm (GA) to account for the latent nonlinear structure of the gene expression in the microarray data. The Davies-Bouldin index is adopted to evaluate the candidate solutions in Isomap and to avoid the classifier dependency problem. Additionally, a probability-based framework is introduced to reduce the possibility of genes being randomly selected by GA. The performance of Iso-GA was evaluated on eight benchmark microarray datasets of cancers. Iso-GA outperformed other benchmarking gene selection methods, leading to good classification accuracy with fewer critical genes selected. CONCLUSIONS: The proposed Iso-GA method can effectively select fewer but critical genes from microarray data to achieve competitive classification performance.

A two-stage hybrid biomarker selection method based on ensemble filter and binary differential evolution incorporating binary African vultures optimization.

BACKGROUND: In the field of genomics and personalized medicine, it is a key issue to find biomarkers directly related to the diagnosis of specific diseases from high-throughput gene microarray data. Feature selection technology can discover biomarkers with disease classification information. RESULTS: We use support vector machines as classifiers and use the five-fold cross-validation average classification accuracy, recall, precision and F1 score as evaluation metrics to evaluate the identified biomarkers. Experimental results show classification accuracy above 0.93, recall above 0.92, precision above 0.91, and F1 score above 0.94 on eight microarray datasets. METHOD: This paper proposes a two-stage hybrid biomarker selection method based on ensemble filter and binary differential evolution incorporating binary African vultures optimization (EF-BDBA), which can effectively reduce the dimension of microarray data and obtain optimal biomarkers. In the first stage, we propose an ensemble filter feature selection method. The method combines an improved fast correlation-based filter algorithm with Fisher score. obviously redundant and irrelevant features can be filtered out to initially reduce the dimensionality of the microarray data. In the second stage, the optimal feature subset is selected using an improved binary differential evolution incorporating an improved binary African vultures optimization algorithm. The African vultures optimization algorithm has excellent global optimization ability. It has not been systematically applied to feature selection problems, especially for gene microarray data. We combine it with a differential evolution algorithm to improve population diversity. CONCLUSION: Compared with traditional feature selection methods and advanced hybrid methods, the proposed method achieves higher classification accuracy and identifies excellent biomarkers while retaining fewer features. The experimental results demonstrate the effectiveness and advancement of our proposed algorithmic model.

Screening and identification of the core immune-related genes and immune cell infiltration in severe burns and sepsis.

Severe burns often have a high mortality rate due to sepsis, but the genetic and immune crosstalk between them remains unclear. In the present study, the GSE77791 and GSE95233 datasets were analysed to identify immune-related differentially expressed genes (DEGs) involved in disease progression in both burns and sepsis. Subsequently, weighted gene coexpression network analysis (WGCNA), gene enrichment analysis, protein-protein interaction (PPI) network construction, immune cell infiltration analysis, core gene identification, coexpression network analysis and clinical correlation analysis were performed. A total of 282 common DEGs associated with burns and sepsis were identified. Kyoto Encyclopedia of Genes and Genomes pathway analysis identified the following enriched pathways in burns and sepsis: metabolic pathways; complement and coagulation cascades; legionellosis; starch and sucrose metabolism; and ferroptosis. Finally, six core DEGs were identified, namely, IL10, RETN, THBS1, FGF13, LCN2 and MMP9. Correlation analysis showed that some core DEGs were significantly associated with simultaneous dysregulation of immune cells. Of these, RETN upregulation was associated with a worse prognosis. The immune-related genes and dysregulated immune cells in severe burns and sepsis provide potential research directions for diagnosis and treatment.

The exploration of shared genes and molecular mechanisms of systemic lupus erythematosus and atherosclerosis.

OBJECTIVE: Despite widespread recognition, the mechanisms underlying the relationship between systemic lupus erythematosus (SLE) and atherosclerosis (AS) are still unclear. Our study aimed to explore the shared genetic signature and molecular mechanisms of SLE and AS using a bioinformatics approach. METHODS: Gene expression profiles of GSE50772 (contains peripheral blood mononuclear cells from 61 SLE patients and 20 normal samples) and GSE100927 (contains 69 AS plaque tissue samples and 35 control samples) were downloaded from the Gene Expression Database (GEO) before the differentially expressed genes were obtained using the "limma" package in R. The differential genes were then subjected to gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis using the DAVID online platform to annotate their functions. The intersection targets of PPI and WGCNA were used as key shared genes for SLE and AS with their diagnostic value as shared genes being verified through ROC curves. Finally, Cytoscape 3.7.2 software was used to construct a miRNA-mRNA network map associated with the shared genes. RESULTS: A total of 246 DEGs were identified, including 189 upregulated genes and 57 downregulated genes, which were mainly enriched in signaling pathways such as TNF signaling pathway, IL-17 signaling pathway, and NF-kB signaling pathway. The molecular basis for the relationship between SLE and AS may be the aforementioned signaling pathways. Following ROC curve validation, the intersection of PPI and WGCNA, as well as AQP9, CCR1, CD83, CXCL1, and FCGR2A, resulted in the identification of 15 shared genes. CONCLUSION: The study provided a new perspective on the common molecular mechanisms between SLE and AS, and the key genes and pathways that were identified as being part of these pathways may offer fresh perspectives and suggestions for further experimental research.

pwrBRIDGE: a user-friendly web application for power and sample size estimation in batch-confounded microarray studies with dependent samples.

Batch effect Reduction of mIcroarray data with Dependent samples usinG Empirical Bayes (BRIDGE) is a recently developed statistical method to address the issue of batch effect correction in batch-confounded microarray studies with dependent samples. The key component of the BRIDGE methodology is the use of samples run as technical replicates in two or more batches, "bridging samples", to inform batch effect correction/attenuation. While previously published results indicate a relationship between the number of bridging samples, M, and the statistical power of downstream statistical testing on the batch-corrected data, there is of yet no formal statistical framework or user-friendly software, for estimating M to achieve a specific statistical power for hypothesis tests conducted on the batch-corrected data. To fill this gap, we developed pwrBRIDGE, a simulation-based approach to estimate the bridging sample size, M, in batch-confounded longitudinal microarray studies. To illustrate the use of pwrBRIDGE, we consider a hypothetical, longitudinal batch-confounded study whose goal is to identify Alzheimer's disease (AD) progression-associated genes from amnestic mild cognitive impairment (aMCI) to AD in human blood after a 5-year follow-up. pwrBRIDGE helps researchers design and plan batch-confounded microarray studies with dependent samples to avoid over- or under-powered studies.

CarbArrayART: a new software tool for carbohydrate microarray data storage, processing, presentation, and reporting.

Glycan microarrays are essential tools in glycobiology and are being widely used for assignment of glycan ligands in diverse glycan recognition systems. We have developed a new software, called Carbohydrate microArray Analysis and Reporting Tool (CarbArrayART), to address the need for a distributable application for glycan microarray data management. The main features of CarbArrayART include: (i) Storage of quantified array data from different array layouts with scan data and array-specific metadata, such as lists of arrayed glycans, array geometry, information on glycan-binding samples, and experimental protocols. (ii) Presentation of microarray data as charts, tables, and heatmaps derived from the average fluorescence intensity values that are calculated based on the imaging scan data and array geometry, as well as filtering and sorting functions according to monosaccharide content and glycan sequences. (iii) Data export for reporting in Word, PDF, and Excel formats, together with metadata that are compliant with the guidelines of MIRAGE (Minimum Information Required for A Glycomics Experiment). CarbArrayART is designed for routine use in recording, storage, and management of any slide-based glycan microarray experiment. In conjunction with the MIRAGE guidelines, CarbArrayART addresses issues that are critical for glycobiology, namely, clarity of data for evaluation of reproducibility and validity.

Batch effect reduction of microarray data with dependent samples using an empirical Bayes approach (BRIDGE).

Batch-effects present challenges in the analysis of high-throughput molecular data and are particularly problematic in longitudinal studies when interest lies in identifying genes/features whose expression changes over time, but time is confounded with batch. While many methods to correct for batch-effects exist, most assume independence across samples; an assumption that is unlikely to hold in longitudinal microarray studies. We propose Batch effect Reduction of mIcroarray data with Dependent samples usinGEmpirical Bayes (BRIDGE), a three-step parametric empirical Bayes approach that leverages technical replicate samples profiled at multiple timepoints/batches, so-called "bridge samples", to inform batch-effect reduction/attenuation in longitudinal microarray studies. Extensive simulation studies and an analysis of a real biological data set were conducted to benchmark the performance of BRIDGE against both ComBat and longitudinalComBat. Our results demonstrate that while all methods perform well in facilitating accurate estimates of time effects, BRIDGE outperforms both ComBat and longitudinal ComBat in the removal of batch-effects in data sets with bridging samples, and perhaps as a result, was observed to have improved statistical power for detecting genes with a time effect. BRIDGE demonstrated competitive performance in batch effect reduction of confounded longitudinal microarray studies, both in simulated and a real data sets, and may serve as a useful preprocessing method for researchers conducting longitudinal microarray studies that include bridging samples.

Differentially expressed EREG and SPP1 are independent prognostic markers in cervical squamous cell carcinoma.

AIMS: Cervical squamous cell carcinoma (SCC) is one of the most frequent malignancies of the female reproductive system. The malignant mechanism of SCC has not been totally clarified. We aimed to discover a list of differentially expressed genes (DEGs) to identify the malignant mechanism of cervical SCC. METHOD: Three expression chips (GSE7803, GSE9750, and GSE64217) were downloaded from gene expression omnibus (GEO) datasets. After standardization, 50 cervical SCC tumor tissues and 33 normal cervical tissues (NCTs) were included for DEGs and clustering analysis. RobustRankAggreg (RRA) algorithm was used to extract the overlapping DEGs. Gene function and signaling pathway analysis was implemented based on Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway databases. Protein-protein interaction (PPI) analysis and prognostic analysis were also carried out to identify the DEGs as prognostic markers for cervical SCC. RESULTS: Totally 100 DEGs were obtained from GSE7803, 319 DEGs from GSE9750, and 1639 DEGs from GSE64217. RRA analysis uncovered 17 upregulated DEGs and 25 downregulated DEGs. GO and KEGG analysis showed DEGs were involved in the mediation of extracellular functions, cell-cell interactions, and cell metabolism. PPI network showed a close interaction among the integrated DEGs. Prognostic analysis showed gene secreted phosphoprotein 1 (SPP1) and epiregulin (EREG) genes were independent prognostic predictors of cervical SCC. CONCLUSION: The gene expression profile was changed in cervical SCC tumor tissues compared to NCTs. SPP1 and EREG were postulated as prognostic markers for cervical SCC, which might be potential targets for clinical therapy of cervical SCC.

Bioinformatics of Differentially Expressed Genes in Phorbol 12-Myristate 13-Acetate-Induced Megakaryocytic Differentiation of K562 Cells by Microarray Analysis.

Megakaryocytes are large hematopoietic cells present in the bone marrow cavity, comprising less than 0.1% of all bone marrow cells. Despite their small number, megakaryocytes play important roles in blood coagulation, inflammatory responses, and platelet production. However, little is known about changes in gene expression during megakaryocyte maturation. Here we identified the genes whose expression was changed during K562 leukemia cell differentiation into megakaryocytes using an Affymetrix GeneChip microarray to determine the multifunctionality of megakaryocytes. K562 cells were differentiated into mature megakaryocytes by treatment for 7 days with phorbol 12-myristate 13-acetate, and a microarray was performed using RNA obtained from both types of cells. The expression of 44,629 genes was compared between K562 cells and mature megakaryocytes, and 954 differentially expressed genes (DEGs) were selected based on a p-value < 0.05 and a fold change >2. The DEGs was further functionally classified using five major megakaryocyte function-associated clusters­inflammatory response, angiogenesis, cell migration, extracellular matrix, and secretion. Furthermore, interaction analysis based on the STRING database was used to generate interactions between the proteins translated from the DEGs. This study provides information on the bioinformatics of the DEGs in mature megakaryocytes after K562 cell differentiation.

Analysis of Genetic Alterations Related to DNA Methylation in Testicular Germ Cell Tumors Based on Data Mining.

Embryonal carcinoma (EC) and seminoma (SE) are both derived from germ cell neoplasia in situ but show big differences in growth patterns and clinical prognosis. Epigenetic regulation may play an important role in the development of EC and SE. This study investigated the DNA methylation-based genetic alterations between EC and SE by analyzing the datasets of mRNA expression and DNA methylation profiling. The datasets were downloaded from the Gene Expression Omnibus database. The differentially expressed genes (DEGs) were identified between EC and SE by limma package in R environment. Gene function enrichment analysis of the DEGs was performed on the DAVID tool, the results of which suggested differences in capability of pluripotency and genomic stability between EC and SE. The minfi package and wANNOVAR tool were used to identify differentially methylated genes. A total of 37 genes were discovered with both mRNA expression and the accordant DNA methylation changes. The findings were verified by the sequencing data from The Cancer Genome Atlas database, and Kaplan-Meier survival analysis was performed. Finally, 5 genes (PRDM1, LMO2, FAM53B, HCN4, and FAM124B) were found that showed both low expression and high methylation in EC, and were significantly associated with relapse-free survival. The findings of methylation-based genetic features between EC and SE might be helpful in studying the role of DNA methylation in cancer development.

Classification of breast cancer using microarray gene expression data: A survey.

Cancer, in particular breast cancer, is considered one of the most common causes of death worldwide according to the world health organization. For this reason, extensive research efforts have been done in the area of accurate and early diagnosis of cancer in order to increase the likelihood of cure. Among the available tools for diagnosing cancer, microarray technology has been proven to be effective. Microarray technology analyzes the expression level of thousands of genes simultaneously. Although the huge number of features or genes in the microarray data may seem advantageous, many of these features are irrelevant or redundant resulting in the deterioration of classification accuracy. To overcome this challenge, feature selection techniques are a mandatory preprocessing step before the classification process. In the paper, the main feature selection and classification techniques introduced in the literature for cancer (particularly breast cancer) are reviewed to improve the microarray-based classification.

Identification of Candidate Genes and Pathways in Nonsegmental Vitiligo Using Integrated Bioinformatics Methods.

BACKGROUND: Nonsegmental vitiligo (NSV) is an acquired depigmentation disorder of unknown origin. Enormous interests focus on finding novel biomarkers and pathways responsible for NSV. METHODS: The gene expression level was obtained by integrating microarray datasets (GSE65127 and GSE75819) from the Gene Expression Omnibus database using the sva R package. Differentially expressed genes (DEGs) between each group were identified by the limma R package. The interaction network was constructed using STRING, and significant modules coupled with hub genes were identified by cytoHubba and molecular complex detection. Pathway analyses were conducted using generally applicable gene set enrichment and further visualized in R environment. RESULTS: A total of 102 DEGs between vitiligo lesional skin and healthy skin, 14 lesion-specific genes, and 29 predisposing genes were identified from the integrated dataset. Except for the anticipated decrease in melanogenesis, three major functional changes were identified, including oxidative phosphorylation, p53, and peroxisome proliferator-activated receptor (PPAR) signaling in lesional skin. PPARG, MUC1, S100A8, and S100A9 were identified as key hub genes involved in the pathogenesis of vitiligo. Besides, upregulation of the T cell receptor signaling pathway was considered to be associated with susceptibility of the skin in NSV patients. CONCLUSION: Our study reveals several potential pathways and related genes involved in NSV using integrated bioinformatics methods. It might provide references for targeted strategies for NSV.

Decreased copy-neutral loss of heterozygosity in African American colorectal cancers.

Despite improvements over the past 20 years, African Americans continue to have the highest incidence and mortality rates of colorectal cancer (CRC) in the United States. While previous studies have found that copy number variations (CNVs) occur at similar frequency in African American and White CRCs, copy-neutral loss of heterozygosity (cnLOH) has not been investigated. In the present study, we used publicly available data from The Cancer Genome Atlas (TCGA) as well as data from an African American CRC cohort, the Chicago Colorectal Cancer Consortium (CCCC), to compare frequencies of CNVs and cnLOH events in CRCs in the two racial groups. Using genotype microarray data, we analyzed large-scale CNV and cnLOH events from 166 microsatellite stable CRCs-31 and 39 African American CRCs from TCGA and the CCCC, respectively, and 96 White CRCs from TCGA. As reported previously, the frequencies of CNVs were similar between African American and White CRCs; however, there was a significantly lower frequency of cnLOH events in African American CRCs compared to White CRCs, even after adjusting for demographic and clinical covariates. Although larger differences for chromosome 18 were observed, a lower frequency of cnLOH events in African American CRCs was observed for nearly all chromosomes. These results suggest that mechanistic differences, including differences in the frequency of cnLOH, could contribute to clinicopathological disparities between African Americans and Whites. Additionally, we observed a previously uncharacterized phenomenon we refer to as small interstitial cnLOH, in which segments of chromosomes from 1 to 5 Mb long were affected by cnLOH.

Microarray data analysis on gene and miRNA expression to identify biomarkers in non-small cell lung cancer.

BACKGROUND: The aim of this study was to gain further investigation of non-small cell lung cancer (NSCLC) tumorigenesis and identify biomarkers for clinical management of patients through comprehensive bioinformatics analysis. METHODS: miRNA and mRNA microarray datasets were downloaded from GEO (Gene Expression Omnibus) database under the accession number GSE102286 and GSE101929, respectively. Genes and miRNAs with differential expression were identified in NSCLC samples compared with controls, respectively. The interaction between differentially expressed genes (DEGs) and differentially expressed miRNAs (DEmiRs) was predicted, followed by functional enrichment analysis, and construction of miRNA-gene regulatory network, protein-protein interaction (PPI) network, and competing endogenous RNA (ceRNA) network. Through comprehensive bioinformatics analysis, we anticipate to find novel therapeutic targets and biomarkers for NSCLC. RESULTS: A total of 123 DEmiRs (5 up- and 118 down-regulated miRNAs) and 924 DEGs (309 up- and 615 down-regulated genes) were identified. These genes and miRNAs were significantly involved in different pathways including adherens junction, relaxin signaling pathway, and axon guidance. Furthermore, hsa-miR-9-5p, has-miR-196a-5p and hsa-miR-31-5p, as well as hsa-miR-1, hsa-miR-218-5p and hsa-miR-135a-5p were shown to have higher degree in the miRNA-gene regulatory network and ceRNA network, respectively. Furthermore, BIRC5 and FGF2, as well as RTKN2 and SLIT3 were hubs in the PPI network and ceRNA network, respectively. CONCLUSION: Several pathways (adherens junction, relaxin signaling pathway, and axon guidance) miRNAs (hsa-miR-9-5p, has-miR-196a-5p, hsa-miR-31-5p, hsa-miR-1, hsa-miR-218-5p and hsa-miR-135a-5p) and genes (BIRC5, FGF2, RTKN2 and SLIT3) may play important roles in the pathogenesis of NSCLC.

Combining gene expression data and prior knowledge for inferring gene regulatory networks via Bayesian networks using structural restrictions.

Gene Regulatory Networks (GRNs) are known as the most adequate instrument to provide a clear insight and understanding of the cellular systems. One of the most successful techniques to reconstruct GRNs using gene expression data is Bayesian networks (BN) which have proven to be an ideal approach for heterogeneous data integration in the learning process. Nevertheless, the incorporation of prior knowledge has been achieved by using prior beliefs or by using networks as a starting point in the search process. In this work, the utilization of different kinds of structural restrictions within algorithms for learning BNs from gene expression data is considered. These restrictions will codify prior knowledge, in such a way that a BN should satisfy them. Therefore, one aim of this work is to make a detailed review on the use of prior knowledge and gene expression data to inferring GRNs from BNs, but the major purpose in this paper is to research whether the structural learning algorithms for BNs from expression data can achieve better outcomes exploiting this prior knowledge with the use of structural restrictions. In the experimental study, it is shown that this new way to incorporate prior knowledge leads us to achieve better reverse-engineered networks.

Analysis of time-course microarray data: Comparison of common tools.

High-throughput time-series data have a special value for studying the dynamism of biological systems. However, the interpretation of such complex data can be challenging. The aim of this study was to compare common algorithms recently developed for the detection of differentially expressed genes in time-course microarray data. Using different measures such as sensitivity, specificity, predictive values, and related signaling pathways, we found that limma, timecourse, and gprege have reasonably good performance for the analysis of datasets in which only test group is followed over time. However, limma has the additional advantage of being able to report significance cut off, making it a more practical tool. In addition, limma and TTCA can be satisfactorily used for datasets with time-series data for all experimental groups. These findings may assist investigators to select appropriate tools for the detection of differentially expressed genes as an initial step in the interpretation of time-course big data.

A hybrid gene selection method based on gene scoring strategy and improved particle swarm optimization.

BACKGROUND: Gene selection is one of the critical steps in the course of the classification of microarray data. Since particle swarm optimization has no complicated evolutionary operators and fewer parameters need to be adjusted, it has been used increasingly as an effective technique for gene selection. Since particle swarm optimization is apt to converge to local minima which lead to premature convergence, some particle swarm optimization based gene selection methods may select non-optimal genes with high probability. To select predictive genes with low redundancy as well as not filtering out key genes is still a challenge. RESULTS: To obtain predictive genes with lower redundancy as well as overcome the deficiencies of traditional particle swarm optimization based gene selection methods, a hybrid gene selection method based on gene scoring strategy and improved particle swarm optimization is proposed in this paper. To select the genes highly related to out samples' classes, a gene scoring strategy based on randomization and extreme learning machine is proposed to filter much irrelevant genes. With the third-level gene pool established by multiple filter strategy, an improved particle swarm optimization is proposed to perform gene selection. In the improved particle swarm optimization, to decrease the likelihood of the premature of the swarm the Metropolis criterion of simulated annealing algorithm is introduced to update the particles, and the half of the swarm are reinitialized when the swarm is trapped into local minima. CONCLUSIONS: Combining the gene scoring strategy with the improved particle swarm optimization, the new method could select functional gene subsets which are significantly sensitive to the samples' classes. With the few discriminative genes selected by the proposed method, extreme learning machine and support vector machine classifiers achieve much high prediction accuracy on several public microarray data, which in turn verifies the efficiency and effectiveness of the proposed gene selection method.

pwrEWAS: a user-friendly tool for comprehensive power estimation for epigenome wide association studies (EWAS).

BACKGROUND: When designing an epigenome-wide association study (EWAS) to investigate the relationship between DNA methylation (DNAm) and some exposure(s) or phenotype(s), it is critically important to assess the sample size needed to detect a hypothesized difference with adequate statistical power. However, the complex and nuanced nature of DNAm data makes direct assessment of statistical power challenging. To circumvent these challenges and to address the outstanding need for a user-friendly interface for EWAS power evaluation, we have developed pwrEWAS. RESULTS: The current implementation of pwrEWAS accommodates power estimation for two-group comparisons of DNAm (e.g. case vs control, exposed vs non-exposed, etc.), where methylation assessment is carried out using the Illumina Human Methylation BeadChip technology. Power is calculated using a semi-parametric simulation-based approach in which DNAm data is randomly generated from beta-distributions using CpG-specific means and variances estimated from one of several different existing DNAm data sets, chosen to cover the most common tissue-types used in EWAS. In addition to specifying the tissue type to be used for DNAm profiling, users are required to specify the sample size, number of differentially methylated CpGs, effect size(s) (Δß), target false discovery rate (FDR) and the number of simulated data sets, and have the option of selecting from several different statistical methods to perform differential methylation analyses. pwrEWAS reports the marginal power, marginal type I error rate, marginal FDR, and false discovery cost (FDC). Here, we demonstrate how pwrEWAS can be applied in practice using a hypothetical EWAS. In addition, we report its computational efficiency across a variety of user settings. CONCLUSION: Both under- and overpowered studies unnecessarily deplete resources and even risk failure of a study. With pwrEWAS, we provide a user-friendly tool to help researchers circumvent these risks and to assist in the design and planning of EWAS. AVAILABILITY: The web interface is written in the R statistical programming language using Shiny (RStudio Inc., 2016) and is available at https://biostats-shinyr.kumc.edu/pwrEWAS/ . The R package for pwrEWAS is publicly available at GitHub ( https://github.com/stefangraw/pwrEWAS ).