Publications by authors named "Tianyi Zang"
Article Synopsis
- The study explores the genetic connections between neurodegenerative diseases, epigenetic aging, and human longevity, using extensive genomic data from a range of diseases and age metrics.
- Results indicated that Alzheimer's disease (AD) is significantly linked to reduced exceptional longevity and has a potential causal relationship with accelerated epigenetic aging.
- The researchers identified shared genetic loci between AD and epigenetic aging, suggesting a complicated interplay of genetics influences across different neurodegenerative diseases, though only AD showed direct causal effects on aging and longevity.
Article Synopsis
- * To address these issues, KGE-UNIT is introduced as a unified framework that integrates knowledge graph embedding and multi-task learning to simultaneously predict drug-target and drug-drug interactions, improving performance even with sparse data.
- * Experiments show that KGE-UNIT significantly outperforms existing methods on various datasets, especially in data-limited situations, demonstrating strong capabilities in predicting other forms of molecular interactions as well.
Glucagon-like peptide 1 (GLP-1) regulates food intake, insulin production, and metabolism. Our recent study demonstrated that pancreatic α-cells-secreted (intraislet) GLP-1 effectively promotes maternal insulin secretion and metabolic adaptation during pregnancy. However, the role of circulating GLP-1 in maternal energy metabolism remains largely unknown.
View Article and Find Full Text PDFThe single-cell proteomics enables the direct quantification of protein abundance at the single-cell resolution, providing valuable insights into cellular phenotypes beyond what can be inferred from transcriptome analysis alone. However, insufficient large-scale integrated databases hinder researchers from accessing and exploring single-cell proteomics, impeding the advancement of this field. To fill this deficiency, we present a comprehensive database, namely Single-cell Proteomic DataBase (SPDB, https://scproteomicsdb.
View Article and Find Full Text PDFGrowing evidence suggests the effect of educational attainment (EA) on Alzheimer's disease (AD), but less is known about the shared genetic architecture between them. Here, leveraging genome-wide association studies (GWAS) for AD (N = 21,982/41,944), EA (N = 1,131,881), cognitive performance (N = 257,828), and intelligence (N = 78,308), we investigated their causal association with the linkage disequilibrium score (LDSC) and Mendelian randomization and their shared loci with the conjunctional false discovery rate (conjFDR), transcriptome-wide association studies (TWAS), and colocalization. We observed significant genetic correlations of EA (r = -0.
View Article and Find Full Text PDFArticle Synopsis
- Predicting how peptides bind to major histocompatibility complex (MHC) is crucial for developing cancer immunotherapies, prompting the creation of the MHCRoBERTa method to improve these predictions using natural language processing (NLP) techniques.
- The MHCRoBERTa model outperforms existing methods, achieving notable increases in the Spearman rank correlation coefficient (SRCC) and the area under the curve (AUC) values, indicating better prediction accuracy.
- Furthermore, the model's ability to visualize token representations showcases its understanding of relevant syntax and semantics, making it a valuable tool for identifying potential neoantigens in cancer treatment, and it is available for public use on GitHub.
Article Synopsis
- Metabolomics plays a crucial role in identifying disease-related metabolites, aiding in disease diagnosis and understanding underlying mechanisms.
- A new model called Disease and Literature Driven Metabolism Prediction Model (DLMPM) was developed to identify potential connections between metabolites and diseases, uncovering 1,406 direct associations and predicting over 119,000 unknown ones.
- DLMPM demonstrates strong performance in predicting metabolic signatures for diseases, achieving an average AUC value of 82.33%, and outperforms previous methods, thus enhancing research on human diseases.
Article Synopsis
- This study highlights the importance of measuring similarity between complex diseases to improve our understanding of disease development and underlying mechanisms.
- It introduces ImpAESim, a new deep-learning method that integrates multiple networks to learn and calculate disease similarity by considering non-coding RNA (ncRNA) and gene interactions.
- The approach reduces biases in disease association calculations, providing more accurate and compact representations of diseases.
Article Synopsis
- Drug-drug interactions (DDIs) are classified into three types: synergistic, antagonistic, and no reaction, and understanding these interactions is crucial in drug development and disease diagnosis.
- The study presents a convolutional neural network (CNN)-based method called CNN-DDI that predicts specific types of drug interactions by analyzing various feature interactions such as drug categories and pathways.
- Results show that utilizing multiple feature types enhances prediction accuracy, making CNN-DDI more effective than existing algorithms in identifying DDIs.
Article Synopsis
- * Pregnant mice showed an increase in α-cell mass and glucagon levels, which correlated with improved glucose metabolism, but when α-cells were absent, glucose tolerance was significantly impaired.
- * GLP-1 receptor activation improved insulin production in the absence of α-cells, suggesting that while α-cells are crucial for metabolic adaptations during pregnancy, glucagon signaling is not essential for maternal glucose management.
Psi-Caller: A Lightweight Short Read-Based Variant Caller With High Speed and Accuracy.
Front Cell Dev Biol
August 2021
With the rapid development of short-read sequencing technologies, many population-scale resequencing studies have been carried out to study the associations between human genome variants and various phenotypes in recent years. Variant calling is one of the core bioinformatics tasks in such studies to comprehensively discover genomic variants in sequenced samples. Many efforts have been made to develop short read-based variant calling approaches; however, state-of-the-art tools are still computationally expensive.
View Article and Find Full Text PDFArticle Synopsis
- Endocrinology studies hormones, their functions as chemical messengers, and the vital roles they play in regulating growth, metabolism, tissue function, and reproduction in mammals through the endocrine system.
- Common endocrine diseases include diabetes mellitus, Grave's disease, and polycystic ovary syndrome, which highlight the importance of understanding their genetic underpinnings to identify pathogenic genes and improve diagnosis and treatment.
- The study introduces a deep learning method called DeepGP, which utilizes advanced neural network techniques to prioritize genes associated with five endocrine diseases, showing promising results that could enhance our understanding of diabetes mellitus and its related disorders.
Summary: Circular consensus sequencing reads are promising for the comprehensive detection of structural variants (SVs). However, alignment-based SV calling pipelines are computationally intensive due to the generation of complete read-alignments and its post-processing. Herein, we propose a SKeleton-based analysis toolkit for Structural Variation detection (SKSV).
View Article and Find Full Text PDFArticle Synopsis
- Single-cell Assay Transposase Accessible Chromatin sequencing (scATAC-seq) is a technique for studying chromatin accessibility across individual cells, but its sparse data poses challenges in identifying cell types compared to single-cell RNA sequencing.
- The authors introduce svmATAC, a support vector machine-based method that improves the identification of cell types by enhancing the signals from scATAC-seq data and filling in missing information through co-accessibility patterns.
- The application of svmATAC to various human immune and blood cell datasets demonstrated its effectiveness and reliability, and the source code for this tool is publicly available on GitHub.
Summary: JavaScript-based Circos libraries have been widely implemented to generate interactive Circos plots in web applications. However, these libraries require either local installation, which requires the compilation of extra libraries, or extra data processing procedures to prepare input and configuration for each track of plot, which limits the utility and capability of integration with powerful R packages. In this report, we present interacCircos, an R package for creating interactive Circos plots through the integration of JavaScript-based libraries.
View Article and Find Full Text PDFInteractions between proteins and small molecule metabolites play vital roles in regulating protein functions and controlling various cellular processes. The activities of metabolic enzymes, transcription factors, transporters and membrane receptors can all be mediated through protein-metabolite interactions (PMIs). Compared with the rich knowledge of protein-protein interactions, little is known about PMIs.
View Article and Find Full Text PDFArticle Synopsis
- Millions struggle with cancer, and while traditional treatments like surgery, radiotherapy, and chemotherapy are used, they can be harmful. A new focus on anticancer peptides (ACPs) offers a safer alternative.
- The authors extracted ACP features based on amino acid sequences and chemical properties, employing deep learning and a method called 'DRACP' to identify true ACPs.
- DRACP showed high accuracy in testing, with performance metrics (AUC and AUPR) above 0.9, proving it more effective than traditional methods in identifying ACPs.
At present, the main diagnostic methods for Alzheimer's disease (AD) are positron emission tomography (PET) scanning of the brain and analysis of cerebrospinal fluid (CSF) sample, but these methods are expensive and harmful to patients. Recently, more researchers focus on diagnosing AD by detecting biomarkers in blood, which is a cheaper and harmless way. Therefore, identifying AD-related proteins in blood can help treatment and diagnosis.
View Article and Find Full Text PDFArticle Synopsis
- Drug side effects are crucial for assessing drug safety, but their frequent occurrence is often due to a lack of clinical understanding and delays in reporting post-approval effects.
- Researchers are actively developing various methods to identify these side effects, categorized into biological experiments, machine learning, text mining, and network approaches.
- The review highlights the strengths and weaknesses of each identification method and suggests future research directions to improve drug safety evaluation.
Article Synopsis
- * Researchers have developed computational methods to improve DTI identification, but many do not integrate drug-protein pair (DPP) associations effectively.
- * A new framework called 'graph convolutional network (GCN)-DTI' was created to include these DPP associations, resulting in better prediction accuracy compared to existing methods.
Article Synopsis
- Peptide-based vaccine development relies on accurately predicting how well peptide ligands bind to MHC I proteins, and current machine learning tools mostly use shallow neural networks for this task.
- Recent research suggests that deep neural networks, specifically convolutional neural networks (CNNs), are more effective in learning from small datasets, which is crucial when only limited peptide data is available for some alleles.
- By incorporating detailed features like sequence order, hydropathy index, and peptide length into a characteristic matrix for input, the proposed CNN-based approach outperforms traditional methods in predicting peptide-MHC binding affinity.
Article Synopsis
- Long-read RNA sequencing alignment is challenging due to high error rates and complex gene structures, necessitating advanced methods.* -
- The proposed method, called deSALT, utilizes a two-pass alignment technique with graph-based skeletons to improve exon inference and refine alignments.* -
- Benchmarks show deSALT outperforms existing methods by producing more accurate full-length alignments and is available on GitHub for access.*
Article Synopsis
- A significant number of long non-coding RNAs (lncRNAs) have been found to be involved in the development of complex human diseases, prompting the need for effective identification methods for disease-related lncRNAs.
- The authors developed a computational model called LncDisAP, which utilizes multiple biological datasets and a collaborative filtering approach to build a functional network of lncRNAs, enabling the identification of potentially disease-related lncRNAs.
- The results demonstrate that LncDisAP successfully predicted novel disease-related lncRNA signatures with an average AUC score of 78.08%, indicating its effectiveness in aiding disease diagnostics and treatment, especially for various cancers.
Background: As the terminal products of cellular regulatory process, functional related metabolites have a close relationship with complex diseases, and are often associated with the same or similar diseases. Therefore, identification of disease related metabolites play a critical role in understanding comprehensively pathogenesis of disease, aiming at improving the clinical medicine. Considering that a large number of metabolic markers of diseases need to be explored, we propose a computational model to identify potential disease-related metabolites based on functional relationships and scores of referred literatures between metabolites.
View Article and Find Full Text PDF