This poster presents a scalable artificial intelligence framework for integrated single-cell analysis of GLP-1 biology across large human gut scRNA-seq datasets. By combining GPU-accelerated preprocessing with RAPIDS single-cell, variational integration using scVI, and foundation model embeddings from scGPT, the framework enables full-resolution analysis of millions of cells without downsampling, preserving rare but biologically critical enteroendocrine populations such as GLP-1-secreting type L EECs. The approach addresses a central challenge in GLP-1 research: integrating heterogeneous datasets across donors, laboratories, and sequencing technologies while maintaining sensitivity to rare cell states. Cross-study harmonization resolves distinct EEC subtypes and supports perturbation-ready modeling of subtype-specific responses to GLP-1 therapeutics, including semaglutide and tirzepatide. This framework enables scalable rare-cell discovery, cross-study integration, and in silico prediction of drug-induced transcriptional programs, providing a foundation for next-generation computational investigation of GLP-1 biology and metabolic therapeutics.
Click here to download a PDF of our poster.
Once you’ve read our poster, if you have any questions or want to explore how we can work with you in a similar capacity, please click here to contact us.
PS – interested in browning more of our posters?
Poster – From Single Cells to Systems: Network Maps for Cluster Identification in Single-Cell Data
Poster – Tumor-Macrophage Hybrid Cells Metastasis
Poster – Raw Reads to Results: Building Next-Generation Bioinformatics Pipelines