Single Cell

Unraveling Liver Fibrosis: The Power of Transformer-Based Single-Cell Analysis in Identifying Key Endothelial and Stellate Cells 

Transformer-based foundation models for single-cell analysis offer unprecedented capabilities in dissecting the complex cellular landscape of the liver cell enables precise identification and annotation of cell subtypes, including liver sinusoidal endothelial cells (LSECs) and stellate cells. This advanced approach has facilitated deeper insights into their pivotal roles in liver fibrosis, illuminating the complex interactions that drive disease progression. The ability to visualize and plot these cell subtypes further enhances our understanding of liver pathology, providing a foundation for targeted therapeutic interventions. Additionally, this analysis can be instrumental in assessing liver toxicity by characterizing cellular responses to toxic agents, offering valuable insights into the mechanisms of liver damage and potential biomarkers for drug safety. At present, this model is being evaluated to gain deeper insights into the role of liver sinusoidal endothelial cells (LSECs) in mediating fibrosis and regeneration, with the aim of developing GenAI-based therapeutic strategies. 

Advanced GenAI and Single-Cell RNA-Seq Platform for Lung Fibrosis Drug Discovery 

The UCEHCA Lung Atlas is a comprehensive single-cell reference atlas of the human lung, integrating data from over 2 million cells across various lung tissues and diseases. This atlas employs GenAI, LLMs, and foundation models for semantic search and analysis at the single-cell level. These advanced tools enable precise identification and characterization of rare cell types and sub-cell types, particularly in idiopathic pulmonary fibrosis (IPF). By leveraging these technologies, researchers can uncover novel insights into disease mechanisms and identify potential therapeutic targets, significantly advancing GenAI-based drug discovery in lung fibrosis. The UCE (Universal Cell Embeddings, a foundation model for cell biology) pretrained (zero-shot) model has been tested and employed to identify key players in lung fibrosis and specific cell types, aiding in the discovery of therapeutic targets. 

Significance of Immune Cell Clustering in Lung Atlas Single-Cell Analysis for Drug Discovery with GenAI 

The clustering of CD1c-positive myeloid dendritic cells, elicited macrophages, and alveolar macrophages in the lung atlas single-cell cluster highlights the coordinated immune response and functional synergy among these cell types. This clustering facilitates efficient pathogen clearance, tissue repair, and immune regulation, underscoring the importance of these cells in maintaining lung health and responding to infections. These three cell types of cluster in the neighborhood due to chemotactic signals and the conducive microenvironment created by alveolar macrophages, which attract and activate elicited macrophages and CD1c-positive myeloid dendritic cells.

Analyzing scRNA-seq data and clustering patterns is crucial for drug discovery as it helps identify key cell types and their interactions within the tissue microenvironment. Understanding these cellular relationships can lead to the development of targeted therapies that modulate specific cell populations, enhancing the efficacy and precision of treatments for various diseases, including cancer and inflammatory conditions.

Creating a knowledge graph of these three cell clusters can further enhance our understanding by visually representing the relationships and interactions between these cell types. This knowledge graph can help identify key signaling pathways, molecular interactions, and potential therapeutic targets. By leveraging GenAI, we can integrate and analyze large-scale data, uncovering hidden patterns and insights that can drive innovative drug discovery and personalized medicine approaches.

The knowledge graph for lung cell clusters is under development and will be released soon to model the relationships of scRNA-seq lung cell cluster data for spatial omics analysis. 

scRNA-Seq Analysis_annotations (liver cells)

-Leiden Plot (left): Cells are colored based on clusters identified by the Leiden algorithm (python analysis), with each color representing a different cluster (0 to 10)

-CellType Plot (right): Cells are colored based on their predicted and identified cell types and annotated accordingly.