Statistical methods for enrichment analysis are important tools to extract biological information from omics experiments. Although these methods have been widely used for the analysis of gene and protein lists, the development of high-throughput technologies for regulatory elements demands dedicated statistical and bioinformatics tools. Here, we present a set of enrichment analysis methods for regulatory elements, including CpG sites, miRNAs, and transcription factors. Statistical significance is determined via a power weighting function for target genes and tested by the Wallenius noncentral hypergeometric distribution model to avoid selection bias. These new methodologies have been applied to the analysis of a set of miRNAs associated with arrhythmia, showing the potential of this tool to extract biological information from a list of regulatory elements. These new methods are available in GeneCodis 4, a web tool able to perform singular and modular enrichment analysis that allows the integration of heterogeneous information.

The increasing use of high-throughput gene expression quantification technologies over the last two decades and the fact that most of the published studies are stored in public databases has triggered an explosion of studies available through public repositories. All this information offers an invaluable resource for reuse to generate new knowledge and scientific findings. In this context, great interest has been focused on meta-analysis methods to integrate and jointly analyze different gene expression datasets. In this work, we describe the main steps in the gene expression meta-analysis, from data preparation to the state-of-the art statistical methods. We also analyze the main types of applications and problems that can be approached in gene expression meta-analysis studies and provide a comparative overview of the available software and bioinformatics tools. Moreover, a practical guide for choosing the most appropriate method in each case is also provided.

Systemic lupus erythematosus (SLE) is a heterogeneous disease with unpredictable patterns of activity. Patients with similar activity levels may have different prognosis and molecular abnormalities. In this study, we aimed to measure the main differences in drug-induced gene expression signatures across SLE patients and to evaluate the potential for clinical data to build a machine learning classifier able to predict the SLE subset for individual patients. SLE transcriptomic data from two cohorts were compared with drug-induced gene signatures from the CLUE database to compute a connectivity score that reflects the capability of a drug to revert the patient signatures. Patient stratification based on drug connectivity scores revealed robust clusters of SLE patients identical to the clusters previously obtained through longitudinal gene expression data, implying that differential treatment depends on the cluster to which patients belongs. The best drug candidates found, mTOR inhibitors or those reducing oxidative stress, showed stronger cluster specificity. We report that drug patterns for reverting disease gene expression follow the cell-specificity of the disease clusters. We used 2 cohorts to train and test a logistic regression model that we employed to classify patients from 3 independent cohorts into the SLE subsets and provide a clinically useful model to predict subset assignment and drug efficacy.

Meta-analysis techniques allow researchers to jointly analyse different studies to determine common effects. In the field of transcriptomics, these methods have gained popularity in recent years due to the increasing number of datasets that are available in public repositories. Despite this, there is a limited number of statistical software packages that implement proper meta-analysis functionalities for this type of data. This article describes DExMA, an R package that provides a set of functions for performing gene expression meta-analyses, from data downloading to results visualization. Additionally, we implemented functions to control the number of missing genes, which can be a major issue when comparing studies generated with different analytical platforms. DExMA is freely available in the Bioconductor repository.

Recent advances in single-cell RNA-Sequencing (scRNA-Seq) technologies have revolutionized our ability to gather molecular insights into different phenotypes at the level of individual cells. The analysis of the resulting data poses significant challenges, and proper statistical methods are required to analyze and extract information from scRNA-Seq datasets. Sample classification based on gene expression data has proven effective and valuable for precision medicine applications. However, standard classification schemas are often not suitable for scRNA-Seq due to their unique characteristics, and new algorithms are required to effectively analyze and classify samples at the single-cell level. Furthermore, existing methods for this purpose have limitations in their usability. Those reasons motivated us to develop singleDeep, an end-to-end pipeline that streamlines the analysis of scRNA-Seq data training deep neural networks, enabling robust prediction and characterization of sample phenotypes. We used singleDeep to make predictions on scRNA-Seq datasets from different conditions, including systemic lupus erythematosus, Alzheimer’s disease and coronavirus disease 2019. Our results demonstrate strong diagnostic performance, validated both internally and externally. Moreover, singleDeep outperformed traditional machine learning methods and alternative single-cell approaches. In addition to prediction accuracy, singleDeep provides valuable insights into cell types and gene importance estimation for phenotypic characterization. This functionality provided additional and valuable information in our use cases. For instance, we corroborated that some interferon signature genes are consistently relevant for autoimmunity across all immune cell types in lupus. On the other hand, we discovered that genes linked to dementia have relevant roles in specific brain cell populations, such as APOE in astrocytes.