Ju Wang

Introduction: Comparing the gene expression profiles of human and non-human primate brains can provide essential insights to understand the structural and functional disparities between these species, as well as the origins of many diseases. In recent years, the availability of genomic and transcriptomic data of the human brain has made it feasible to explore the evolutionary features of the brain-expressed genes at the molecular level on a large scale. Analyses of human and non-human primate brain cell-type-specific expressed genes with elevated evolutionary rates can uncover the genetic and functional diversity intrinsic to the human brain. Methods: In this study, we try to explore the genetic diversity of the human brain by analyzing genes with divergent evolutionary rates in human and macaque brains, as well as in the human lineage. Results: We identified 450 genes with high evolutionary rates between human and macaque, and 534 genes between human populations. Topological characteristics showed that these genes had lower protein-protein interaction network node degree. We also found that pathways related to neurological diseases, immune, metabolism, and signaling transduction were enriched in the genes of high evolutionary rates between human and macaque. We further identified 97 and 137 potential genes with differential evolution features between and within species. Discussion: Our results reveal the genetic diversity of evolutionarily variable specific-expressed genes in the human brain and provide complementary findings for human brain evolution at the transcriptional level.

Genetic interactions (GIs) underpin the functional connectivity of genes and pathways, and are important for dissecting genotype–phenotype relationships and identifying therapeutic targets for diseases. However, the scale of the human genome restricts systematic experimental interrogation of GIs. Existing computational tools focus on predicting synthetic lethality (SL) and synthetic viability (SV), the two primary forms of GIs, yet their accuracy and biological interpretability are compromised by inadequate modeling of the molecular mechanisms behind positive and negative interactions, as well as the limitation of negative samples. To overcome these challenges, we developed Genetic Interaction Graph Convolutional Network (GiGCN), a signed network modeling framework for the joint identification of gene pairs with SL and SV. We built a high-confidence signed genetic network by integrating verified GIs, and non-interacting gene pairs, together with gene semantic similarity derived from biological processes. By leveraging disentangled subspace decomposition, this framework separately models distinct functional dimensions within gene networks, enabling robust representation of context-dependent regulatory relationships and accurate discrimination of SL and SV events. Benchmark experiments demonstrate that GiGCN outperforms state-of-the-art approaches (area under receiver operating-characteristic curve: 0.978, and area under precision–recall curve: 0.944). Further analyses reveal biologically meaningful insights, including known and novel SL interactions centered on the oncogene MYC Proto-Oncogene (MYC), as well as SV interactions linked to autophagy and mitophagy pathways. This study provides a robust and interpretable network-based strategy for systematically exploring GIs. The GiGCN framework not only improves the precision of SL and SV prediction, but also offers mechanistic insights into gene functional relationships, thereby supporting the discovery of actionable therapeutic targets for cancer and other human diseases.

Transcription factors (TFs) bind to specific sequences in the genome to regulate gene expression and specify cell states. TF binding sites (TFBSs) are cell type-specific, which can be attributed to epigenomic contexts. Comprehensive profiling of TFBSs across various cell types through experimental approaches is neither practical nor cost-friendly. Accurately identifying cell type-specific TFBSs through computational approaches remains challenging. Here, we develop EpiXFormer, a novel transformer-based neural network for cell type-specific TFBS prediction. EpiXFormer achieves exceptional performance in predicting binding sites of DNA-binding proteins (DBPs) across a diverse collection of cell types. It models the effects of proximal and distal epigenomic information on DBP binding and learns the identified motifs of the examined TFs and their potential co-occurring proteins. Moreover, we demonstrate that EpiXFormer can infer pioneer factors during cell type transition and delineate the cell type-specific regulatory functions of TFs. Overall, EpiXFormer enables cell type-specific TFBS prediction in the examined cell lines and is readily applied to other cell types of interest. It provides a robust, scalable framework for characterizing and interpreting multimodal genomic data.

Background: Alzheimer's disease (AD) is a severe neurodegenerative disorder, yet its molecular mechanisms remain incompletely understood. It is known that the joint action of a number of genetic and other factors is involved in the pathogenesis of this disorder. Objective: In the past years, extensive research has focused on identifying novel AD predictors and genetic markers. However, our understanding of the molecular features of AD is still incomplete, and it is essential to discover novel genes and their interaction involved in the etiology and development of AD. Methods: Here, we developed GRESA (Gene RElationship Sequence Analyzer), to predict novel genes related to disease by a combination of machine learning algorithm Skip-gram and Monte Carlo Tree Search (MCTS). First, we extracted gene association information contained in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database and generated gene relationship sequences. Then, Skip-gram neural network was trained and combined with MCTS to predict genes potentially associated with diseases like AD. Results: The performance of GRESA was evaluated on breast cancer and gastric cancer. We further constructed the gene network underlying AD pathogenesis via GRESA, by which 164 potential gene candidates were predicted and their molecular function and biological features were investigated. Conclusions: Results from this study provided insights for understanding the molecular feature underlying AD. As a useful systematic method, our approach can also be applied to construct gene networks for other complex diseases.

Neurological disorders (e.g., Alzheimer’s disease and Parkinson’s disease) and mental disorders (e.g., depression and anxiety), pose huge challenges to global public health. The pathogenesis of these diseases can usually be attributed to many factors, such as genetic, environmental and socioeconomic status, which make the diagnosis and treatment of the diseases difficult. As research on the diseases advances, so does the body of medical data. The accumulation of such data provides unique opportunities for the basic and clinical study of these diseases, but the vast and diverse nature of the data also make it difficult for physicians and researchers to precisely extract the information and utilize it in their work. A powerful tool to extract the necessary knowledge from large amounts of data is knowledge graph (KG). KG, as an organized form of information, has great potential for the study neurological and mental disorders when it is paired with big data and deep learning technologies. In this study, we reviewed the application of KGs in common neurological and mental disorders in recent years. We also discussed the current state of medical knowledge graphs, highlighting the obstacles and constraints that still need to be overcome.

Alzheimer’s disease (AD) is the most prevalent neurodegenerative disease that is characterized by memory loss and cognitive impairment. Evidence shows that depression is a common co-occurrence in AD patients, and major depressive disorder (MDD) is considered a risk factor for AD. The crosstalk between the biological procedures related to the two disorders makes it very difficult to treat the comorbid conditions caused by them. Considering the common pathophysiological mechanisms underlying AD and MDD, antidepressant drugs may have beneficial therapeutic effects against their concurrence. In this study, we aimed to explore the potential drug candidates for the prevention and treatment of the comorbidity of AD and MDD. First, we screened the potential drugs for treating MDD by evaluating the distances of drug targets to MDD-related genes on the human protein–protein interaction network (PPIN) via a network-based algorithm. Then, the drugs were further screened to identify those that may be effective for AD treatment by analyzing their affinities with tau protein and Aβ42 peptide via molecular docking. Furthermore, the most stable binding modes were identified via molecular dynamics simulations, and the regulatory effects of drug candidates on genes involved in the pathogenesis of AD and MDD were analyzed. A total of 506 MDD-related genes were retrieved, and 831 drug candidates for MDD treatment were screened via the network-based approach. The results from molecular docking and molecular dynamics simulations indicated dihydroergotamine had the lowest binding affinity with tau protein and bromocriptine could form the most stable binding mode with Aβ42 peptide. Further analyses found that both dihydroergotamine and bromocriptine could regulate the expression of genes involved in the pathogenesis of AD and/or MDD in the brain. The exact mechanisms of the two drugs in treating AD and MDD, as well as their comorbidity, are still unclear, and further exploration is needed to evaluate their roles and mechanisms, both in vitro and in vivo. This study revealed that dihydroergotamine and bromocriptine may be the potential drug candidates for the treatment of the comorbidity of AD and MDD, and the therapeutic effects may be achieved by inhibiting the accumulation and aggregation of Aβ42 and tau protein and regulating the expression of disease-related genes in the brain.

The nicotinic acetylcholine receptors (nAChR) are complexes that are composed of subunits evolved from common ancestor. Although the subunits are similar in sequences and structures, their molecular function varies significantly. Therefore, detecting the molecular sites specific to each subunit is important to understand the property of the subunits and the receptors formed by them. As we know, the molecular sites critical to the structure and molecular function of a protein family usually are conserved in evolution, and those specific to each member of the family are often closely related to its structural and functional specificity. In this study, we analyzed the sequence specific sites in nAChR subunits by adopting the evolutionary trace method and the two-state model, and explored the relationship between structure and function in combination with the spatial location of the sites. The results showed that the detected sites in α7 subunit were closely related to ligand binding and conformational changes of the energetic coupling pathway. The conserved sites tended to be distributed in the interior of the spatial structure of protein molecules, and the sites potentially related to new functions were distributed on the surface of the spatial structure. In summary, our results could be helpful to understanding the molecular features related to the function specificity and diversity of the nAChR subunits.