Xinlei Wang

Rank aggregation, which aims to combine multiple ordinal rankings into a single, reliable ranking, is widely studied in diverse fields. In academic and professional settings, peer- and self-evaluations play a pivotal role and offer insights into both individual performance and the competence of rankers. However, subjective biases can lead to unreliable or unfair outcomes, especially in a small sample setting. Despite the broad applications of rank aggregation, existing methods often fail to address nuanced biases and the special structure of such data. This paper introduces BayeSPeer, a novel Bayesian hierarchical model designed to jointly estimate latent performance scores and self-assessment biases in peer- and self-ranking data. Through comprehensive simulation studies, BayeSPeer demonstrates superior performance in rank recovery and top-k identification under various levels of sample size, ranking quality, and bias magnitudes. Real-world applications in classroom peer evaluations also highlight the model’s interpretability and practical utility. Our findings showcase the advantages of incorporating structured bias correction rather than omitting self-evaluation, offering a statistically principled and transparent approach to bias-aware ranking in educational and other evaluation contexts.

Summary: In this paper, we address the normal mean inference problem, which involves testing multiple means of normal random variables with heteroscedastic variances. Most existing empirical Bayes methods for this setting are developed under restrictive assumptions, such as the scaled inverse‐chi‐squared prior for variances and unimodality for the non‐null mean distribution. However, when either of these assumptions is violated, these methods often fail to control the false discovery rate (FDR) at the target level or suffer from a substantial loss of power. To overcome these limitations, we propose a new empirical Bayes method, gg‐Mix, which assumes only independence between the normal means and variances, without imposing any structural restrictions on their distributions. We thoroughly evaluate the FDR control and power of gg‐Mix through extensive numerical studies and demonstrate its superior performance compared to existing methods. Finally, we apply gg‐Mix to three real data examples to further illustrate the practical advantages of our approach.

Predicting drug sensitivity and selecting informative biomarkers are fundamental challenges in precision oncology. These tasks are complicated by the ultrahigh-dimensional nature of omics data, where the number of covariates p far exceeds the sample size n. While variable selection methods are widely used, traditional high-dimensional regression techniques often rely on the assumption of Gaussian noise, which is frequently violated in real-world omics datasets. To overcome these limitations, we propose a Bayesian empirical likelihood approach (BEL_HD) for variable selection and prediction in a semiparametric framework. BEL_HD avoids strict distributional assumptions by utilizing estimating equations and imposes joint regularization on both regression coefficients and the Lagrange multipliers to promote sparsity. We also introduce an efficient active set MCMC algorithm that enables scalable inference in ultrahigh-dimensional spaces. Through simulations and real-data applications, including a case study on leukemia drug sensitivity, BEL_HD demonstrates superior predictive performance and selects biologically interpretable features. Our method combines the flexibility of empirical likelihood with the inferential benefits of Bayesian modeling, offering a robust and extensible tool for high-dimensional biomedical research.

In meta-analyses of continuous outcomes, the sample mean and standard deviation (SD) are essential for synthesizing effect sizes across studies. However, clinical studies frequently report alternative summary statistics, such as the median, quartiles, and range. To enable inclusion of such studies, various methods have been proposed to estimate the sample mean and SD from these reported summaries. We propose the Bayesian Order Statistics-based Estimator (BOSE), which leverages the joint likelihood of observed order statistics together with weakly informative priors to obtain the full posterior distribution for the mean and SD without relying on computationally intensive iterative procedures such as Markov chain Monte Carlo algorithms. Our numerical studies demonstrate that BOSE performs competitively with existing approaches in estimating the mean, while achieving superior performance for estimating the SD across all evaluated scenarios, particularly in small-sample settings. Under non-normal distributions including skewed, heavy-tailed, and bimodal settings with mild or moderate deviations from normality, BOSE remains robust and stable, whereas methods specifically designed for skewed distributions may become unstable or even inapplicable. Beyond point estimation, BOSE naturally provides empirically validated posterior credible intervals, enabling researchers to formally quantify uncertainty for study-level estimates and make reliable, evidence-based decisions in meta-analytic research synthesis. A publicly accessible web application implementing BOSE and competing methods is also provided to facilitate practical use in meta-analytic research.

With the rise of large-scale genomic studies, large gene lists targeting important diseases are increasingly common. While evaluating each study individually gives valuable insights on specific samples and study designs, the wealth of available evidence in the literature calls for robust and efficient meta-analytic methods. Crucially, the diverse assumptions and experimental protocols underlying different studies require a flexible but rigorous method for aggregation. To address these issues, we propose BiGER, a Bayesian rank aggregation method for the inference of latent global rankings. Unlike existing methods in the field, BiGER accommodates mixed gene lists with top-ranked and top-unranked genes as well as bottom-tied and missing genes, by design. Using a Bayesian hierarchical framework combined with variational inference, BiGER efficiently aggregates large-scale gene lists, consistently achieving state-of-the-art accuracy, while providing valuable insights into source-specific reliability for researchers. Through both simulated and real datasets, we show that BiGER is a useful tool for reliable meta-analysis in genomic studies.

Transcriptional regulators (TRs) are essential proteins that regulate gene transcription, and their dysregulation defines the oncogenic transcriptional regulatory programs that drive tumor-specific gene expression. Existing pan-cancer resources summarize these programs by aggregating signals across individual datasets, neglecting the context-specific features that capture the diversity of transcriptional regulation underlying cancer heterogeneity. By developing a variational Bayesian hierarchical model named eTRex ( e pigenomics-based T ranscriptional R egulator ex plorer) and applying it to 4,819 cancer-related ATAC-seq datasets, we provide a comprehensive pan-cancer atlas of functional TR profiles that preserve the context-specific features of each dataset. This study reveals both common regulators across diverse malignancies and those with highly specific roles. We extensively validated these findings using independent CRISPR/Cas9 screening, mutation, and transcriptomic datasets. Collectively, this pan-cancer atlas of functional TR profiles represents a comprehensive, biologically interpretable resource for uncovering transcriptional regulatory programs, identifying biomarkers, prioritizing therapeutic targets in oncology, and is freely accessible through an interactive web portal.

Imaging-based spatially resolved transcriptomics (SRT) technologies, such as 10X Xenium, MERSCOPE, and CosMx, have revolutionized our ability to study gene expression within the spatial context of tissues at single-cell resolution. The acquisition of such data relies heavily on cell segmentation algorithms, which often produce imperfect boundaries, leading to transcript misassignment. These misassignments can significantly affect downstream analyses, including cell type identification, differential expression analysis, cell-cell communication, and RNA localization. We present MisTIC (Missegmented Transcript Inference Correction), a variational Bayesian model designed to correct transcript misassignment errors without requiring resegmentation. In benchmarking analyses using synthetic data with simulated transcript misassignment, MisTIC demonstrated high sensitivity and specificity in removing misassigned transcripts. In real data applications, MisTIC effectively enhances cell type identification, reduces ambiguity in differential expression analysis, and improves the detection of cell-cell communication. Furthermore, RNA localization analysis based on data corrected by MisTIC revealed that, in T cells located near cancer-associated fibroblasts compared to those farther away, genes involved in T cell activation, inflammation, and cytotoxicity were depleted from cytoplasmic regions despite not being differentially expressed between the two T cell subsets. In conclusion, MisTIC is a powerful tool for correcting transcript misassignment in SRT data. It not only improves the accuracy of routine analyses but also enables novel investigations that provide deeper insights into the dynamics of gene expression.