Biology subjects

Li, N. T.

Publications and source records attributed to Li, N. T..

6 recordsLinked to original sources

bioRxiv2025

Application of Large Language Models for Annotating Genes into Reactome Pathways

Reactome is the most comprehensive, open source, open access biological pathway knowledgebase, widely used in the research community. To ensure the highest quality of its content, human pathway data in Reactome is manually curated. However, manual curation is labor-intensive, time-consuming, and increasingly difficult to keep up with the ever-growing biomedical literature. Large language model (LLM)-driven artificial intelligence (AI) technologies are transforming many fields, including bioinformatics resource development. Applying LLM/AI technologies in Reactome may offer a powerful way to scale curation and consolidate pathway-related data into a single resource. This manuscript describes the first stage of our attempt to adopt LLM/AI technologies for Reactome manual curation. We developed an LLM workflow that can assist curators in adding new genes to existing pathways and refining the functional annotations of existing ones. The workflow predicts pathways in which genes are likely to function, identifies PubMed-indexed literature that may support these predictions, generates text summaries describing potential molecular mechanisms, and extracts functional relationships among biological entities from full-text PDF papers. To validate the workflow output, we used a computational approach based on semantic similarity between LLM workflow-generated summaries and Reactome manual annotations. The results show significant enrichment of high-similarity matches. Manual evaluation of 19 genes indicated that more than half of the outputs are useful for supporting curation. Based on these results, we developed an enhanced workflow that incorporates protein-protein interaction data, facilitating Reactomes reaction-based annotation. In summary, our initial adoption of LLM/AI technologies produced encouraging results and provides a practical framework for integrating AI-assisted methods into Reactomes curation pipeline. The strategies described here may be broadly applicable to community knowledgebases in general.

bioinformatics↗

bioRxiv2025

In-House Manufacturing of 3D Culture Chips via Vacuum Thermoforming for Enhanced Imaging Applications

Engineered 3D in vitro cancer models, particularly those that facilitate image-based readouts capable of distinguishing the behavior of different cell populations, have become crucial tools in the discovery process. One such model, 96-GLAnCE (Gels for Live Analysis of Compartmentalized Environments), allows for longitudinal imaging of tumor cell dynamics and therapy response. However, the widespread adoption of 96-GLAnCE has been limited by the need for expensive, specialized fabrication equipment. To overcome this challenge, we have optimized a desktop vacuum thermoforming technique for the in-house production of 96- GLAnCE bottom chips using thin polystyrene films. This optimization has led to the reliable and consistent fabrication of devices. Notably, using thin polystyrene films reduces the overall thickness of the chips, enabling high-magnification imaging for studying primary tumor cell phenotypes in 3D with single-cell and subcellular resolution. Our thermoformed devices offer a flexible, cost-effective solution for addressing a wide range of biological questions across various time scales.

bioengineering↗

bioRxiv2024

A cryopreservation strategy for myoblast storage in paper-based scaffolds for inter-laboratory studies of skeletal muscle health

Three-dimensional tissue-engineered models are poised to facilitate understanding of skeletal muscle pathophysiology and identify novel therapeutic agents to improve muscle health. Adopting these culture models within the broader biology community is a challenge as many models involve complex methodologies and significant investments of time and resources to optimize manufacturing protocols. To alleviate this barrier, we developed a protocol with commercially available reagents to cryopreserve myoblasts in a 96-well compatible format that allows tissues to be transferred to users without expertise in 2D or 3D skeletal muscle cell culture. We validate that myoblasts encapsulated in a hydrogel and cryopreserved in paper-based scaffolds maintain cell viability, differentiation, and function via acetylcholine-induced transient calcium responses. Furthermore, we demonstrate successful shipping of myoblasts cryopreserved in paper-based scaffolds to intra-provincial and international collaborators who successfully thawed, cultured, and used the 3D muscle tissues. Finally, we confirm the application of our method to study muscle endogenous repair by seeding freshly isolated skeletal muscle stem cells to cryopreserved then differentiated and injured tissues, demonstrating expected responses to a known stimulator of muscle stem cell self-renewal, p38/{beta} MAPKi. Altogether, our 3D myoblast cryopreservation protocol offers broadened access of a complex skeletal muscle tissue model to the research community.

bioengineering↗

bioRxiv2022

An automation workflow for high-throughput manufacturing and analysis of scaffold-supported 3D tissue arrays

The success rate of bringing novel cancer therapies to the clinic remains extremely low due to the lack of relevant pre-clinical culture models that capture the complexity of human tumours. Patient-derived organoids have emerged as a useful tool to model patient and tumour heterogeneity to begin addressing this need. Scaling these complex culture models while enabling stratified analysis of different cellular sub-populations remains a challenge, however. One strategy to enable higher throughput organoid cultures that also enables easy image-based analysis is the Scaffold-supported Platform for Organoid-based Tissues (SPOT) platform. SPOT allows the generation of flat, thin and dimensionally-defined microtissues in both 96- and 384-well plate footprints and is compatible with tumour organoid culture and downstream image-based readouts. SPOT manufacturing is currently a manual process however, limiting the use of SPOT to perform larger-scale screening. In this study, we integrate and optimize an automation approach to generate tumour-mimetic 3D engineered microtissues in SPOT using a liquid handler, and show comparable within-sample and between-sample variation as the standard manual manufacturing process. Furthermore, we develop a liquid handler-supported whole-cell extraction protocol and as a proof-of-value demonstration, we generate 3D complex tissues containing different proportions of tumour and stromal cells and perform single-cell-based end-point analysis to demonstrate the impact of co-culture on the tumour cell population specifically. We also demonstrate we can incorporate primary patient-derived organoids into the pipeline to capture patient-level tumour heterogeneity. We envision that this automated workflow integrated with 96/384-SPOT and multiple cell types and patient-derived organoid models will provide opportunities for future applications in high-throughput screening for novel personalized therapeutic targets. This pipeline also allows the user to assess dynamic cell responses using high-content longitudinal imaging or downstream single-cell-based analyses.

bioengineering↗

bioRxiv2022

An off-the-shelf multi-well scaffold-supported platform for tumour organoid-based tissues

Complex 3D bioengineered tumour models provide the opportunity to better capture the heterogeneity of patient tumours. Patient-derived organoids are emerging as a useful tool to study tumour heterogeneity and variation in patient responses. Organoid cultures typically require a 3D microenvironment that can be manufactured easily to facilitate screening. Here we set out to create a high-throughput, "off-the-shelf" platform which permits the generation of organoid-containing microtissues for standard phenotypic bioassays and image-based readings. To achieve this, we developed the Scaffold-supported Platform for Organoid-based Tissues (SPOT) platform. SPOT is a 3D gel-embedded in vitro platform that can be produced in a 96- or 384-well plate format and enables the generation of flat, thin and dimensionally-defined microgels. SPOT has high potential for adoption due to its reproducible manufacturing methodology, compatibility with existing instrumentation, and reduced within-sample and between-sample variation, which can pose challenges to both data analysis and interpretation. Using SPOT we generate cultures from patient derived pancreatic ductal adenocarcinoma organoids and assess the cellular response to standard-of-care chemotherapeutic compounds, demonstrating our platforms usability for drug screening. We envision 96/384-SPOT will provide a useful tool to assess drug sensitivity of patient-derived organoids and easily integrate into the drug discovery pipeline.

bioengineering↗

bioRxiv2022

A three-dimensional human adipocyte model of fatty acid-induced obesity

Obesity prevalence has reached pandemic proportions, leaving individuals at high risk for the development of diseases such as cancer and type 2 diabetes. In obesity, to accommodate excess lipid storage, adipocytes become hypertrophic, which is associated with an increased pro-inflammatory cytokine secretion and dysfunction of metabolic processes such as insulin signaling and lipolysis. Targeting adipocyte dysfunction is an important strategy to prevent the development of obesity-associated disease. However, it is unclear how accurately animal models reflect human biology, and the long-term culture of human hypertrophic adipocytes in an in vitro 2D monolayer is challenging due to the buoyant nature of adipocytes. Here we describe the development of a human 3D in vitro disease model that recapitulates hallmarks of obese adipocyte dysfunction. First, human primary adipose-derived stromal cells are embedded in hydrogel, and infiltrated into a thin cellulose scaffold. The thin microtissue profile allows for efficient assembly and image-based analysis. After adipocyte differentiation, the scaffold is stimulated with oleic or palmitic acid to mimic caloric overload. Using functional assays, we demonstrated that this treatment induced important obese adipocyte characteristics such as a larger lipid droplet size, increased basal lipolysis, insulin resistance and activation of macrophages through adipocyte-conditioned media. This 3D disease model mimics physiologically relevant hallmarks of obese adipocytes, to enable investigations into the mechanisms by which dysfunctional adipocytes contribute to disease.

bioengineering↗