A5033831822- Cancer Genomics and Diagnostics
- Bioinformatics and Genomic Networks
- Glioma Diagnosis and Treatment
- Gene expression and cancer classification
- Gene Regulatory Network Analysis
- Computational Drug Discovery Methods
- Prostate Cancer Treatment and Research
- Radiomics and Machine Learning in Medical Imaging
- Monoclonal and Polyclonal Antibodies Research
- vaccines and immunoinformatics approaches
- Immune cells in cancer
- Genomics and Phylogenetic Studies
- Glycosylation and Glycoproteins Research
- Cancer Cells and Metastasis
- Molecular Biology Techniques and Applications
- Acoustic Wave Phenomena Research
- Neuroinflammation and Neurodegeneration Mechanisms
- Microbial Metabolic Engineering and Bioproduction
- Vascular Anomalies and Treatments
- Ferroptosis and cancer prognosis
- Connexins and lens biology
- Machine Learning in Materials Science
- Machine Learning in Bioinformatics
- Moyamoya disease diagnosis and treatment
- Hearing, Cochlea, Tinnitus, Genetics
Tencent (China)
2022-2024
University of Hong Kong
2018-2023
Hong Kong University of Science and Technology
2018-2023
Max Delbrück Center
2023
Chinese University of Hong Kong
2020
Nano and Advanced Materials Institute
2018
Purdue University West Lafayette
2014-2017
BackgroundHepatocellular carcinoma (HCC) often presents with multiple nodules within the liver, limited effective interventions. The high genetic heterogeneity of HCC might be major cause treatment failure. We aimed to characterize genomic heterogeneity, infer clonal evolution, investigate RNA expression pattern and explore tumour immune microenvironment profile multifocal HCC.Patients methodsWhole-exome sequencing were carried out in 34 tumours 6 adjacent normal liver tissue samples from...
The role of brain immune compartments in glioma evolution remains elusive. We profile cells microenvironment and the matched peripheral blood from 11 patients. Glioblastoma exhibits specific infiltration blood-originated monocytes expressing epidermal growth factor receptor (EGFR) ligands EREG AREG, coined as tumor-associated (TAMo). TAMo is mutually exclusive with EGFR alterations (p = 0.019), while co-occurring mesenchymal subtype 4.7 × 10−7) marking worse prognosis 0.004 0.032 two...
Diffusion-based network models are widely used for protein function prediction using data and have been shown to outperform neighborhood-based module-based methods. Recent studies that integrating the hierarchical structure of Gene Ontology (GO) dramatically improves accuracy. However, previous methods usually either GO hierarchy refine results multiple classifiers, or flattened into a function-function similarity kernel. No study has taken account together with as two-layer model.We first...
Abstract Accurate prediction of antibody-antigen complex structures holds significant potential for advancing biomedical research and the design therapeutic antibodies. Currently, structure protein monomers has achieved considerable success, promising progress been made in extending this achievement to complexes. However, despite these advancements, fast accurate remains a challenging unresolved issue. Existing end-to-end methods, which rely on homology templates, exhibit sub-optimal...
Rationale: Brain arteriovenous malformations (bAVMs) are abnormal entanglement of blood vessels in brain, with direct connections from arteries to veins, lacking functional capillary bed. Although several somatic mutations were reported, the molecular mechanism and genetic disposition bAVM remain poorly understood. Objective: We aim identify transcriptional anomalies critical pathways lesions explore their association key de novo germline variants patients. Methods Results: established a...
Abstract Accurate prediction of antibody structures is critical in analyzing the function antibodies, thus enabling rational design antibodies. However, existing structure methods often only formulate backbone atoms and rely on additional tools for side-chain conformation prediction. In this work, we propose a fully end-to-end architecture simultaneous conformations, namely tFold-Ab. Pre-trained language models are adopted fast by avoiding time-consuming search sequence homologs. The model...
Clonal evolution drives cancer progression and therapeutic resistance. Recent studies have revealed divergent longitudinal trajectories in gliomas, but early molecular features steering posttreatment remain unclear. Here, we collected sequencing clinical data of initial-recurrent tumor pairs from 544 adult diffuse gliomas performed multivariate analysis to identify predictors three glioma subtypes. We found that CDKN2A deletion at initial diagnosis preceded necrosis microvascular...
Abstract Metastatic cancer is associated with poor patient prognosis but its spatiotemporal behavior remains unpredictable at early stage. Here we develop MetaNet, a computational framework that integrates clinical and sequencing data from 32,176 primary metastatic cases, to assess risks of tumors. MetaNet achieves high accuracy in distinguishing the metastasis breast prostate cancers. From prediction, identify Metastasis-Featuring Primary (MFP) tumors, subset tumors genomic features...
Subnetwork detection is often used with differential expression analysis to identify modules or pathways associated a disease condition. Many computational methods are available for subnetwork analysis. Here, we compare the results of eight methods: simulated annealing-based jActiveModules, greedy search-based DEGAS, BioNet, NetBox, ClustEx, OptDis, and NetWalker. These represent distinctly different strategies among most widely used. Each these was analyze gene data consisting paired tumor...
The complex pattern of cancer evolution poses a huge challenge to precision oncology. Longitudinal sequencing tumor samples allows us monitor the dynamics mutations that occurred during this clonal process. Here, we present versatile toolbox, namely CELLO (Cancer EvoLution for LOngitudinal data), accompanied with step‐by‐step tutorial, exemplify how profile, analyze and visualize dynamic change somatic mutational landscape using longitudinal genomic data. Moreover, customize hypermutation...
The mechanotransduction (MT) complex in auditory hair cells converts the mechanical stimulation of sound waves into neural signals. Recently, MT has been suggested to contain at least four distinct integral membrane proteins: protocadherin 15 (PCDH15), transmembrane channel-like protein 1 (TMC1), lipoma HMGIC fusion partner-like 5 (LHFPL5), and inner ear (TMIE). However, composition, function, regulation MT-complex proteins remain incompletely investigated. Here, we report previously...
Abstract Clonal evolution drives cancer progression and therapeutic resistance 1-2 . Recent longitudinal analyses revealed divergent clonal dynamics in adult diffuse gliomas 3–11 However, the early genomic epigenomic factors that steer post-treatment molecular trajectories remain unknown. To track evolutionary predictors, we analyzed sequencing clinical data of matched initial-recurrent tumor pairs from 511 glioma patients. Using machine learning developed methods capable predicting grade...
Protein fluxes provide a more refined notion of protein abundance than raw counts alone by considering potential channels based on interaction networks. We propose novel method to estimate in network using linear programming model the framework flux balance analysis. When we combine this with protein-centric measure, inspired egocentric analysis sociology, discover that proteins encoded hypermutated genes colon cancer have substantially higher alterations cells quantities alone. These remain...
Diffusion-based network models are widely used for protein function prediction using data and have been shown to outperform neighborhood- module-based methods. Recent studies that integrating the hierarchical structure of Gene Ontology (GO) dramatically improves accuracy. However, previous methods usually either GO hierarchy refine results multiple classifiers, or flattened into a function-function similarity kernel. No study has taken account together with as two-layer model. We first...
Proteins govern a wide range of biological systems. Evaluating the changes in protein properties upon mutation is fundamental application design, where modeling 3D structure principal task for AI-driven computational approaches. Existing deep learning (DL) approaches represent as geometric graph and simplify to different degrees, thereby failing capture low-level atom patterns high-level amino acid simultaneously. In addition, limited training samples with ground truth labels structures...
Abstract Glioblastoma (GBM) is a lethal brain tumor with limited therapeutic options. Therapeutic resistance arises from the collaboration among heterogeneous and highly plastic cellular entities states conserved across GBM patients microenvironment.Tumor cells adapt in response to current standard of care infiltration innate immune cells, recurrently acquiring mesenchymal state that drives resistance.Moreover, blood-brain barrier (BBB) poses challenges effectiveness bioavailability approved...
Abstract Metastasis leads to most cancer deaths, but its spatiotemporal behavior remains unpredictable at early stage. Here, we developed MetaNet, a computational framework that integrates clinical and sequencing data from 32,176 primary metastatic cases, assess risks of tumors. MetaNet achieved high accuracy in distinguishing the metastasis breast prostate cancers. From prediction, identified Metastasis-Featuring Primary (MFP) tumors, subset tumors with genomic features enriched metastasis,...