Biology subjects

Lau, M. C.

Publications and source records attributed to Lau, M. C..

6 recordsLinked to original sources

bioRxiv2026

Spatial patterns of APOBEC mutagenesis in the tumour microenvironment of Asian breast cancer

A germline deletion of the coding region at the APOBEC3B (A3B) gene that is common in Asian and Oceanic populations has been found to be correlated with increased APOBEC mutagenesis, immune activation, and tumour heterogeneity in breast cancer. The mechanism by which A3B deletion and APOBEC mutagenesis induces immune activation in Asian breast cancers is still unclear, but has been thought to involve increased neoantigen burden. In this study, we investigated the mechanisms by which A3B deletion and APOBEC mutagenesis may lead to immune activation using Stereo-seq spatial transcriptomics of Malaysian triple-negative breast cancer (TNBC) tumours with and without the A3B deletion to examine the spatial expression patterns of APOBEC3A/B as well the spatial patterns of APOBEC-associated mutations and neoantigens. Unexpectedly, we found only limited associations between A3B deletion, APOBEC-associated neoantigens, and immune activation. Instead, we found strong spatial associations between mitochondrial APOBEC-associations mutations, the cGAS-STING pathway, and the NF-{kappa}B pathway. Our study thus supports a new model whereby APOBEC-mediated activation of the immune system does not occur via neoantigen presentation, but instead occurs via mitochondrial hypermutation that drives cGAS-STING activation of the NF-{kappa}B pathway. Our study may provide new insights into the spatial landscape of Asian breast tumours and the effect of germline A3B deletion and APOBEC mutagenesis on the tumour immune microenvironment.

cancer biology↗

bioRxiv2025

A Domain-Specific Language Model Approach for Identifying Immunogenic Epitopes from Publicly Available Data: EpitopeMiner

Personalized cancer vaccines represent a transformative approach in immunotherapy, leveraging tumor-specific antigens, such as neoantigens, to stimulate durable and targeted immune responses. However, identifying immunogenic neoantigen epitopes through conventional approaches - relying on in silico predictions followed by experimental validation - remains a significant challenge due to limitations of computational tools and scattering of experiment data. To address these challenges, we introduce EpitopeMiner, a domain-specific large language model (LLM) enhanced with a Retrieval-Augmented Generation (RAG) framework, specifically designed to identify immunogenic epitopes predicted by existing computational tools. EpitopeMiner utilizes a custom database of MHC-I-associated epitope-related literature to provide domain-specific knowledge, enhancing the precision and relevance of LLM-generated responses. EpitopeMiner offers three key features: First, the ability to identify epitopes with similar sequences and potentially similar immunogenic effects, which is especially valuable for neoantigens that are patient-specific and rarely found in public datasets. Second, it supports multiple epitope searches with structured outputs, enhancing scalability. Third, it provides original text chunks and paper identifiers, significantly simplifying validation and further exploration of the retrieved knowledge. Applying EpitopeMiner to well-characterized MHC class I epitopes demonstrates its ability to consistently retrieve relevant papers, efficiently providing targeted insights on T-cell response and immunogenicity, outperforming a commercial AI-powered literature tool. Notably, when applied to lymphoma patient-derived neoantigens, EpitopeMiner successfully retrieves immune response information for a few similar epitopes, despite operating with a relatively smaller database compared to the benchmark tool. All in all, EpitopeMiner bridges the gap between computational prediction and experimental validation, providing a scalable solution for extracting knowledge from public data, fostering cross-study synergies, and accelerating the development of personalized cancer vaccines.

cancer biology↗

bioRxiv2023

Comparing virtual staining artificial intelligence model performance on same-tissue versus serial-tissue sections based on CD3+ T-cell ground truth

Immunophenotyping via multi-marker assays significantly contributes to patient selection, therapeutic monitoring, biomarker discovery, and personalized treatments. Despite its potential, the multiplex immunofluorescence (mIF) technique faces adoption challenges due to technical and financial constraints. Alternatively, hematoxylin and eosin (H&E)-based prediction models of cell phenotypes can provide crucial insights into tumor-immune cell interactions and advance immunotherapy. Current methods mostly rely on manually annotated cell label ground truths, with limitations including high variability and substantial labor costs. To mitigate these issues, researchers are increasingly turning to digitized cell-level data for accurate in-situ cell type prediction. Typically, immunohistochemical (IHC) staining is applied to a tissue section serial to one stained with H&E. However, this method may introduce distortions and tissue section shifts, challenging the assumption of consistent cellular locations. Conversely, mIF overcomes these limitations by allowing for mIF and H&E staining on the same tissue section. Importantly, the multiplexing capability of mIF allows for a thorough analysis of the tumor microenvironment by quantifying multiple cell markers within the same tissue section. In this study, we introduce a Pix2Pix generative adversarial network (P2P-GAN)-based virtual staining model, using CD3+ T-cells in lung cancer as a proof-of-concept. Using an independent CD3 IHC-stained lung cohort, we demonstrate that the model trained with cell label ground-truth from the same tissue section as H&E staining performed significantly better in both CD3+ and CD3- T-cell prediction. Moreover, the model also displayed prognostic significance on a public lung cohort, demonstrating its potential clinical utility. Notably, our proposed P2P-GAN virtual staining model facilitates image-to-image translation, enabling further spatial analysis of the predicted immune cells, deepening our understanding of tumor-immune interactions, and propelling advancements in personalized immunotherapy. This concept holds potential for the prediction of other cell phenotypes, including CD4+, CD8+, and CD20+ cells.

immunology↗

bioRxiv2022

Probabilistic embedding and clustering with alignment for spatial transcriptomics data integration with PRECAST

Spatially resolved transcriptomics involves a set of emerging technologies that enable the transcriptomic profiling of tissues with the physical location of expressions. Although a variety of methods have been developed for data integration, most of them are for single-cell RNA-seq datasets without consideration of spatial information. Thus, methods that can integrate spatial transcriptomics data from multiple tissue slides, possibly from multiple individuals, are needed. Here, we present PRECAST, a data integration method for multiple spatial transcriptomics datasets with complex batch effects and/or biological effects between slides. PRECAST unifies spatial factor analysis simultaneously with spatial clustering and embedding alignment, while requiring only partially shared cell/domain clusters across datasets. Using both simulated and four real datasets, we show improved cell/domain detection with outstanding visualization, and the estimated aligned embeddings and cell/domain labels facilitate many downstream analyses. We demonstrate that PRECAST is computationally scalable and applicable to spatial transcriptomics datasets from different platforms.

genomics↗

bioRxiv2020

Single cell RNA-seq reveals immunosuppressive gastric stem-like cancer cells as a poor prognostic factor

ObjectiveGastric cancer (GC) tumors are highly heterogenous with different subpopulations of epithelial cells. We employed single cell RNA sequencing (scRNA-seq) to dissect the heterogeneity and identified subpopulations of cancer cells with stem-like properties. We further investigated their resistance to oxaliplatin chemotherapy and their contribution to gastric cancer outcome. DesignWe performed scRNA-seq on FACS sorted epithelial and immune cells from paired samples of GC tumors and normal adjacent tissues. We identified two epithelial subpopulations (STMN1+IQGAP3+ and STMN1+IQGAP3-) with stem-like properties. We characterized and compared them to known healthy gastric stem cell populations. We also cultivated GC derived organoids to study the chemoresistance of similarly marked populations. Lastly, we employed immunohistochemistry (IHC) staining to ascertain the predicted immunosuppressive interactions. ResultsThe STMN1+IQGAP3+ subpopulation showed a higher tumor mutation burden, upregulated proliferative pathways and transcriptomically resembled proliferative healthy gastric isthmus stem cells. The STMN1+IQGAP3- subpopulation were comparatively quiescent and transcriptomically resembled enteroendocrine cells. Both transcriptomic signatures were associated with worse mortality than other epithelial subpopulations with the quiescent being associated with the poorest patient survival. GC tissue derived organoids were dominated by STMN1+IQGAP3+ cells but the STMN1+IQGAP3- compartment was more resistant to chemotherapy. We also verified the likely suppression of CD8 T cell cytotoxicity by STMN1+IQGAP3+ cells through the NECTIN2/TIGIT interaction. ConclusionsCancer cells with stem-like characteristics are associated with poor survival through chemoresistance and immunosuppression. Reactivating the immune system through checkpoint blockade is an opportunity to eliminate these cells. What is already known on this topicMultiple gastric stem cell populations have been identified and linked to tumor initiation in rodent-based studies. However, none of them have been conclusively proven in human tumors. Isolating and characterizing tumor cells with stem-like properties will help shed light on their possible origin and possible mitigation strategies. What this study addsHere we identified two sets of stem-like gastric cancer cells that are associated with poorer patient prognosis. One set is highly proliferative and exhibits oxaliplatin susceptibility. It also engages in immunosuppressive interactions such as NECTIN2/TIGIT. The other set is quiescent and highly resistant to oxaliplatin. How this study might affect research, practice or policyThe transcriptome signatures of the identified stem-like cells can aid in patient prognosis and identify patients who can benefit from checkpoint blockade therapy to reactivate their immune response towards gastric cancer cells.

immunology↗

bioRxiv2020

Tumor-Immune Partitioning and Clustering (TIPC) algorithm reveals distinct signatures of tumor-immune cell interactions within the tumor microenvironment

Growing evidence supports the importance of understanding tumor-immune spatial relationship in the tumor microenvironment in order to achieve precision cancer therapy. However, existing methods, based on oversimplistic cell-to-cell proximity, are largely confounded by immune cell density and are ineffective in capturing tumor-immune spatial patterns. Here we developed a novel computational algorithm, termed Tumor-Immune Partitioning and Clustering (TIPC), to offer an effective solution for spatially informed tumor subtyping. Our method could measure the extent of immune cell partitioning between tumor epithelial and stromal areas as well as the degree of immune cell clustering. Using a U.S. nation-wide colorectal cancer database, we showed that TIPC could determine tumor subtypes with unique tumor-immune spatial patterns that were significantly associated with patient survival and key tumor molecular features. We also demonstrated that TIPC was robust to parameter settings and readily applicable to different immune cell types. The capability of TIPC in delineating clinically relevant patient subtypes that encapsulate tumor-immune spatial relationship, immune density, and tumor morphology is expected to shed light on underlying immune mechanisms. Hence, TIPC can be a useful bioinformatics tool for effective characterization of the spatial composition of the tumor-immune microenvironment to inform precision immunotherapy.

cancer biology↗