Biology subjects

Le, D.-H.

Publications and source records attributed to Le, D.-H..

6 recordsLinked to original sources

bioRxiv2026

PAG-Agent: a biologist-oriented research assistant for context-aware pathway-level analysis and interpretation

Pathway analysis is a critical step for translating gene-level omics results into biological mechanisms, yet existing workflows often leave researchers with long lists of statistically significant pathways that are difficult to interpret, validate, and connect to experimental context. We developed PAG-Agent, a biologist-oriented virtual research assistant that integrates pathway-level statistical analysis, context-aware biological interpretation, literature-supported reasoning, and scientific writing support within a unified workflow. PAG-Agent supports bulk and single-cell transcriptomic data and enables users to perform data preprocessing, differential expression analysis, pathway analysis, pathway-level consensus analysis, and pathway-level meta-analysis through click-based and chat-based interactions. Unlike conventional pathway-analysis tools that analyze gene sets largely in isolation, PAG-Agent incorporates experimental conditions and research objectives to prioritize biologically relevant pathways and generate interpretable hypotheses. The system also provides gene and pathway annotation, citation retrieval, visualization, and writing refinement functions. In Alzheimers disease case studies using three transcriptomic datasets, PAG-Agent consistently identified neurodegeneration-related pathways across multiple analysis methods and datasets. In citation-retrieval benchmarking, PAG-Agent outperformed six competing LLMs across five common literature-support scenarios, demonstrating improved ability to provide contextually relevant and valid references. Overall, PAG-Agent lowers technical barriers for pathway-level analysis and helps researchers move from transcriptomic data to biologically grounded interpretation, hypothesis generation, and scientific communication.

bioinformatics↗

bioRxiv2025

VN1K: a genome graph-based and function-driven multi-omics and phenomics resource for the Vietnamese population

Vietnam, the 16th most populated nation, remains profoundly underrepresented in global genomic databases. Here, we present VN1K, a first-ever comprehensive and well-curated resource of multi-omics data with a wide-range of phenotypic information of 1,011 unrelated Vietnamese individuals. High-depth short-read whole-genome sequencing data were generated for all samples along with various - omic data, including microarray, long-read whole-genome sequencing, and RNA sequencing. Using a high-sensitivity variant detection pipeline, which included a pangenome graph reference and a deep-learning framework, we identified nearly 40 million variants of which 8.5 million are novel with nearly 900 thousand short insertions/deletions and 39 thousand structural variants. Specifically, VN1K featured a first-ever whole-genome methylation profile based on long read sequencing. A genotype imputation panel was also created with the highest accuracy on the Vietnamese population. Variants with significantly different allele frequencies in the Vietnamese population compared to others were found to be functionally significant, especially in genes associated with immune diseases (HLA-B, KIR3DL3, KIR2DL1, KIR2DL4) or drug responses (CYP2C19, CYP2D6, VKORC1, CYP2B6). We were also able to map various loci related to hepatitis B virus infection as well as six disease traits, including triglyceride levels, LDL-C, serum glucose levels, HbA1c, and levels of two liver enzymes (ALT and AST). VN1K dataset is accessible via genome.vinbigdata.org, an integrated platform with both linear and graph-based genome browser for facilitating data exploration, research, and applications in precision medicine.

genomics↗

bioRxiv2022

A biphasic Deep Semi-supervised framework for Suptype Classification and biomarker discovery

To take full advantage of the unprecedented development of -omics technologies and generate further biological insights into human disease, it is a pressing need to develop novel computational methods for integrative analysis of multi-omics data. Here we proposed a biphasic Deep Semi-supervised multi-omics integration framework for Subtype Classification and biomarker discovery, DeepSSC. In phase 1, each denoising autoencoder was used to extract a compact representation for each -omics data, and then they were concatenated and put into a feed-forward neural network for subtype classification. In phase 2, our Biomarker Gene Identification procedure leveraged that neural network classifier to render subtype-specific important biomarkers. We also validated our given results on independent dataset. We demonstrated that DeepSSC exhibited better performance over other state-of-the-art techniques concerning classification tasks. As a result, DeepSSC successfully detected well-known biomarkers and hinted at novel candidates from different -omics data types related to the investigated biomedical problems.

bioinformatics↗

bioRxiv2021

overlappingCGM: Automatic detection and analysis of overlapping co-expressed gene modules

BackgroundWhen it comes to the co-expressed gene module detection, its typical challenges consist of overlap between identified modules and local co-expression in a subset of biological samples. The nature of module detection is the use of unsupervised clustering approaches and algorithms. Those methods are advanced undoubtedly, but the selection of a certain clustering method for sample- and gene-clustering tasks is separate, in which the latter task is often more complicated. ResultsThis study presented an R-package, Overlapping CoExpressed gene Module (oCEM), armed with the decomposition methods to solve the challenges above. We also developed a novel auxiliary statistical approach to select the optimal number of principal components using a permutation procedure. We showed that oCEM outperformed state-of-the-art techniques in the ability to detect biologically relevant modules additionally. ConclusionsoCEM helped non-technical users easily perform complicated statistical analyses and then gain robust results. oCEM and its applications, along with example data, were freely provided at https://github.com/huynguyen250896/oCEM.

bioinformatics↗

bioRxiv2020

Graph convolutional networks for drug response prediction

BackgroundDrug response prediction is an important problem in computational personalized medicine. Many machine-learning-based methods, especially deep learning-based ones, have been proposed for this task. However, these methods often represent the drugs as strings, which are not a natural way to depict molecules. Also, interpretation (e.g., what are the mutation or copy number aberration contributing to the drug response) has not been considered thoroughly. MethodsIn this study, we propose a novel method, GraphDRP, based on graph convolutional network for the problem. In GraphDRP, drugs were represented in molecular graphs directly capturing the bonds among atoms, meanwhile cell lines were depicted as binary vectors of genomic aberrations. Representative features of drugs and cell lines were learned by convolution layers, then combined to represent for each drug-cell line pair. Finally, the response value of each drug-cell line pair was predicted by a fully-connected neural network. Four variants of graph convolutional networks were used for learning the features of drugs. ResultsWe found that GraphDRP outperforms tCNNS in all performance measures for all experiments. Also, through saliency maps of the resulting GraphDRP models, we discovered the contribution of the genomic aberrations to the responses. ConclusionRepresenting drugs as graphs can improve the performance of drug response prediction. Availability of data and materialsData and source code can be downloaded at https://github.com/hauldhut/GraphDRP.

bioinformatics↗

bioRxiv2020

Unveiling Prognostics Biomarkers of Tyrosine Metabolism Reprogramming in Liver Cancer by Cross-platform Gene Expression Analyses

Tyrosine is mainly degraded in the liver by a series of enzymatic reactions. Abnormal expression of the tyrosine catabolic enzyme tyrosine aminotransferase (TAT) has been reported in patients with hepatocellular carcinoma (HCC). Despite this, aberration in tyrosine metabolism has not been investigated in cancer development. In this work, we conduct comprehensive cross-platform study to obtain foundation for discoveries of potential therapeutics and preventative biomarkers of HCC. We explore data from The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO), Gene Expression Profiling Interactive Analysis (GEPIA), Oncomine and Kaplan Meier plotter (KM plotter) and performed integrated analyses to evaluate the clinical significance and prognostic values of the tyrosine catabolic genes in HCC. We find that five tyrosine catabolic enzymes are downregulated in HCC compared to normal liver at mRNA and protein level. Moreover, low expression of these enzymes correlates with poorer survival in patients with HCC. Notably, we identify pathways and upstream regulators that might involve in tyrosine catabolic reprogramming and further drive HCC development. In total, our results underscore tyrosine metabolism alteration in HCC and lay foundation for incorporating these pathway components in therapeutics and preventative strategies.

cancer biology↗