Integrating Genetic and Gene Co-expression Analysis Identifies Gene Networks Involved in Alcohol and Stress Responses

Applications of gene pair methods in clinical research: advancing precision medicine

The rapid evolution of high-throughput sequencing technologies has revolutionized biomedical research, producing vast amounts of gene expression data that hold immense potential for biological discovery and clinical applications. Effectively mining these large-scale, high-dimensional data is crucial for facilitating disease detection, subtype differentiation, and understanding the molecular mechanisms underlying disease progression. However, the conventional paradigm of single-gene profiling, measuring absolute expression levels of individual genes, faces critical limitations in clinical implementation. These include vulnerability to batch effects and platform-dependent normalization requirements. In contrast, emerging approaches analyzing relative expression relationships between gene pairs demonstrate unique advantages. By focusing on binary comparisons of two genes' expression magnitudes, these methods inherently normalize experimental variations while capturing biologically stable interaction patterns. In this review, we systematically evaluate gene pair-based analytical frameworks. We classify eleven computational approaches into two fundamental categories: expression value-based methods quantifying differential expression patterns, and rank-based methods exploiting transcriptional ordering relationships. To bridge methodological development with practical implementation, we establish a reproducible analytical pipeline incorporating feature selection, classifier construction, and model evaluation modules using real-world benchmark datasets from pulmonary tuberculosis studies. These findings position gene pair analysis as a transformative paradigm for mining high-dimensional omics data, with direct implications for precision biomarker discovery and mechanistic studies of disease progression.

Transcriptomic coordination at hepatic steatosis indicates robust immune cell engagement prior to inflammation

BACKGROUND

Deregulation in lipid metabolism leads to the onset of hepatic steatosis while at subsequent stages of disease development, the induction of inflammation, marks the transition of steatosis to non-alcoholic steatohepatitis. While differential gene expression unveils individual genes that are deregulated at different stages of disease development, how the whole transcriptome is deregulated in steatosis remains unclear.

METHODS

Using outbred deer mice fed with high fat as a model, we assessed the correlation of each transcript with every other transcript in the transcriptome. The onset of steatosis in the liver was also evaluated histologically.

RESULTS

Our results indicate that transcriptional reprogramming directing immune cell engagement proceeds robustly, even in the absence of histologically detectable steatosis, following administration of high fat diet. In the liver transcriptomes of animals with steatosis, a preference for the engagement of regulators of T cell activation and myeloid leukocyte differentiation was also recorded as opposed to the steatosis-free livers at which non-specific lymphocytic activation was seen. As compared to controls, in the animals with steatosis, transcriptome was subjected to more widespread reorganization while in the animals without steatosis, reorganization was less extensive. Comparison of the steatosis and non-steatosis livers showed high retention of coordination suggesting that diet supersedes pathology in shaping the transcriptome's profile.

CONCLUSIONS

This highly versatile strategy suggests that the molecular changes inducing inflammation proceed robustly even before any evidence of steatohepatitis is recorded, either histologically or by differential expression analysis.

Resilience, plasticity and robustness in gene expression during aging in the brain of outbred deer mice

BACKGROUND

Genes that belong to the same network are frequently co-expressed, but collectively, how the coordination of the whole transcriptome is perturbed during aging remains unclear. To explore this, we calculated the correlation of each gene in the transcriptome with every other, in the brain of young and older outbred deer mice (P. leucopus and P. maniculatus).

RESULTS

In about 25 % of the genes, coordination was inversed during aging. Gene Ontology analysis in both species, for the genes that exhibited inverse transcriptomic coordination during aging pointed to alterations in the perception of smell, a known impairment occurring during aging. In P. leucopus, alterations in genes related to cholesterol metabolism were also identified. Among the genes that exhibited the most pronounced inversion in their coordination profiles during aging was THBS4, that encodes for thrombospondin-4, a protein that was recently identified as rejuvenation factor in mice. Relatively to its breadth, abolishment of coordination was more prominent in the long-living P. leucopus than in P. maniculatus but in the latter, the intensity of de-coordination was higher.

CONCLUSIONS

There sults suggest that aging is associated with more stringent retention of expression profiles for some genes and more abrupt changes in others, while more subtle but widespread changes in gene expression appear protective. Our findings shed light in the mode of the transcriptional changes occurring in the brain during aging and suggest that strategies aiming to broader but more modest changes in gene expression may be preferrable to correct aging-associated deregulation in gene expression.

Deep learning of gene relationships from single cell time-course expression data

Time-course gene-expression data have been widely used to infer regulatory and signaling relationships between genes. Most of the widely used methods for such analysis were developed for bulk expression data. Single cell RNA-Seq (scRNA-Seq) data offer several advantages including the large number of expression profiles available and the ability to focus on individual cells rather than averages. However, the data also raise new computational challenges. Using a novel encoding for scRNA-Seq expression data, we develop deep learning methods for interaction prediction from time-course data. Our methods use a supervised framework which represents the data as 3D tensor and train convolutional and recurrent neural networks for predicting interactions. We tested our time-course deep learning (TDL) models on five different time-series scRNA-Seq datasets. As we show, TDL can accurately identify causal and regulatory gene-gene interactions and can also be used to assign new function to genes. TDL improves on prior methods for the above tasks and can be generally applied to new time-series scRNA-Seq data.

Network preservation reveals shared and unique biological processes associated with chronic alcohol abuse in NAc and PFC

Chronic alcohol abuse has been linked to the disruption of executive function and allostatic conditioning of reward response dysregulation in the mesocorticolimbic pathway (MCL). Here, we analyzed genome-wide mRNA and miRNA expression from matched cases with alcohol dependence (AD) and controls (n = 35) via gene network analysis to identify unique and shared biological processes dysregulated in the prefrontal cortex (PFC) and nucleus accumbens (NAc). We further investigated potential mRNA/miRNA interactions at the network and individual gene expression levels to identify the neurobiological mechanisms underlying AD in the brain. By using genotyped and imputed SNP data, we identified expression quantitative trait loci (eQTL) uncovering potential genetic regulatory elements for gene networks associated with AD. At a Bonferroni corrected p≤0.05, we identified significant mRNA (NAc = 6; PFC = 3) and miRNA (NAc = 3; PFC = 2) AD modules. The gene-set enrichment analyses revealed modules preserved between PFC and NAc to be enriched for immune response processes, whereas genes involved in cellular morphogenesis/localization and cilia-based cell projection were enriched in NAc modules only. At a Bonferroni corrected p≤0.05, we identified significant mRNA/miRNA network module correlations (NAc = 6; PFC = 4), which at an individual transcript level implicated miR-449a/b as potential regulators for cellular morphogenesis/localization in NAc. Finally, we identified eQTLs (NAc: mRNA = 37, miRNA = 9; PFC: mRNA = 17, miRNA = 16) which potentially mediate alcohol's effect in a brain region-specific manner. Our study highlights the neurotoxic effects of chronic alcohol abuse as well as brain region specific molecular changes that may impact the development of alcohol addiction.

Analysis of the genes controlling three quantitative traits in three diverse plant species reveals the molecular basis of quantitative traits

Most traits of agricultural importance are quantitative traits controlled by numerous genes. However, it remains unclear about the molecular mechanisms underpinning quantitative traits. Here, we report the molecular characteristics of the genes controlling three quantitative traits randomly selected from three diverse plant species, including ginsenoside biosynthesis in ginseng (Panax ginseng C.A. Meyer), fiber length in cotton (Gossypium hirsutum L. and G. barbadense L.) and grain yield in maize (Zea mays L.). We found that a vast majority of the genes controlling a quantitative trait were significantly more likely spliced into multiple transcripts while they expressed. Nevertheless, only one to four, but not all, of the transcripts spliced from each of the genes were significantly correlated with the phenotype of the trait. The genes controlling a quantitative trait were multiple times more likely to form a co-expression network than other genes expressed in an organ. The network varied substantially among genotypes of a species and was associated with their phenotypes. These findings indicate that the genes controlling a quantitative trait are more likely pleiotropic and functionally correlated, thus providing new insights into the molecular basis underpinning quantitative traits and knowledge necessary to develop technologies for efficient manipulation of quantitative traits.

Integrating transcriptomic network reconstruction and eQTL analyses reveals mechanistic connections between genomic architecture and Brassica rapa development

Plant developmental dynamics can be heritable, genetically correlated with fitness and yield, and undergo selection. Therefore, characterizing the mechanistic connections between the genetic architecture governing plant development and the resulting ontogenetic dynamics of plants in field settings is critically important for agricultural production and evolutionary ecology. We use hierarchical Bayesian Function-Valued Trait (FVT) models to estimate Brassica rapa growth curves throughout ontogeny, across two treatments, and in two growing seasons. We find genetic variation for plasticity of growth rates and final sizes, but not the inflection point (transition from accelerating to decelerating growth) of growth curves. There are trade-offs between growth rate and duration, indicating that selection for maximum yields at early harvest dates may come at the expense of late harvest yields and vice versa. We generate eigengene modules and determine which are co-expressed with FVT traits using a Weighted Gene Co-expression Analysis. Independently, we seed a Mutual Rank co-expression network model with FVT traits to identify specific genes and gene networks related to FVT. GO-analyses of eigengene modules indicate roles for actin/cytoskeletal genes, herbivore resistance/wounding responses, and cell division, while MR networks demonstrate a close association between metabolic regulation and plant growth. We determine that combining FVT Quantitative Trait Loci (QTL) and MR genes/WGCNA eigengene expression profiles better characterizes phenotypic variation than any single data type (i.e. QTL, gene, or eigengene alone). Our network analysis allows us to employ a targeted eQTL analysis, which we use to identify regulatory hotspots for FVT. We examine cis vs. trans eQTL that mechanistically link FVT QTL with structural trait variation. Colocalization of FVT, gene, and eigengene eQTL provide strong evidence for candidate genes influencing plant height. The study is the first to explore eQTL for FVT, and specifically do so in agroecologically relevant field settings.

We estimate the developmental dynamics of plant growth using mathematical functions to fit continuous functions to discrete plant height data collected throughout growth, and we use the parameters defining these mathematical functions as data. We identify genomic regions controlling plant growth and filter a novel transcriptomic data set using network reconstruction models to identify the genes and eigengenes associated with plant height. We combine these genomic and transcriptomic data to predict variation in plant height, and we use quantitative genetics to mechanistically connect plant genetics, transcriptomics, and development. Our approach demonstrates two powerful methods for the type of data reduction (FVT modeling and gene expression network reconstruction for targeted eQTL analyses) and data integration that will be necessary for driving forward the field of genetics in the post-genomic era. To the best of our knowledge, we are the first to apply these techniques to continuous models of plant development, and the first to do so in agroecologically relevant field settings.

Studying alcohol use disorder using Drosophila melanogaster in the era of 'Big Data'

Our understanding of the networks of genes and protein functions involved in Alcohol Use Disorder (AUD) remains incomplete, as do the mechanisms by which these networks lead to AUD phenotypes. The fruit fly (Drosophila melanogaster) is an efficient model for functional and mechanistic characterization of the genes involved in alcohol behavior. The fly offers many advantages as a model organism for investigating the molecular and cellular mechanisms of alcohol-related behaviors, and for understanding the underlying neural circuitry driving behaviors, such as locomotor stimulation, sedation, tolerance, and appetitive (reward) learning and memory. Fly researchers are able to use an extensive variety of tools for functional characterization of gene products. To understand how the fly can guide our understanding of AUD in the era of Big Data we will explore these tools, and review some of the gene networks identified in the fly through their use, including chromatin-remodeling, glial, cellular stress, and innate immunity genes. These networks hold great potential as translational drug targets, making it prudent to conduct further research into how these gene mechanisms are involved in alcohol behavior.

Evaluation of Sirtuin-3 probe quality and co-expressed genes using literature cohesion

BACKGROUND

Gene co-expression studies can provide important insights into molecular and cellular signaling pathways. The GeneNetwork database is a unique resource for co-expression analysis using data from a variety of tissues across genetically distinct inbred mice. However, extraction of biologically meaningful co-expressed gene sets is challenging due to variability in microarray platforms, probe quality, normalization methods, and confounding biological factors. In this study, we tested whether literature derived functional cohesion could be used as an objective metric in lieu of 'ground truth' to evaluate the quality of probes and microarray datasets.

RESULTS

We examined Sirtuin-3 (Sirt3) co-expressed gene sets extracted from either liver or brain tissues of BXD recombinant inbred mice in the GeneNetwork database. Depending on the microarray platform, there were as many as 26 probes that targeted different regions of Sirt3 primary transcript. Co-expressed gene sets (ranging from 100-1000 genes) associated with each Sirt3 probe were evaluated using the previously developed literature-derived cohesion p-value (LPv) and benchmarked against 'gold standards' derived from proteomic studies or Gene Ontology classifications. We found that the maximal F-measure was obtained at an average window size of 535 genes. Using set size of 500 genes, the Pearson correlations between LPv and F-measure as well as between LPv and mitochondrial gene enrichment p-values were 0.90 and 0.93, respectively. Importantly, we found that the LPv approach can distinguish high quality Sirt3 probes. Analysis of the most functionally cohesive Sirt3 co-expressed gene set revealed core metabolic pathways that were shared between hippocampus and liver as well as distinct pathways which were unique to each tissue. These results are consistent with other studies that suggest Sirt3 is a key metabolic regulator and has distinct functions in energy-producing vs. energy-demanding tissues.

CONCLUSIONS

Our results provide proof-of-concept that literature cohesion analysis is useful for evaluating the quality of probes and microarray datasets, particularly when experimentally derived gold standards are unavailable. Our approach would enable researchers to rapidly identify biologically meaningful co-expressed gene sets and facilitate discovery from high throughput genomic data.