A-T.49: Benchmarking Cell-Level, Pseudobulk, and Metacell Differential Gene Expression Analyses for Single Nucleus Transcriptomics Data – A Real-World Use Case
Author
Keywords
pharmaceutical, neuroscience, metacell, DGEA, bnechmark, real-world application
[sponser-meet-now-chat][/sponser-meet-now-chat]
Single-nucleus DGE analyses suffer from extreme p-values when treating cells as independent replicates. Bulking mitigates pseudo replication but discards single cell information (heterogeneity, distributions). We evaluate metacells - inspired by high dimensional weighted gene correlation network analyses - to preserve within population structure while stabilizing variance.
We compared cell-level, metacell, and pseudobulk DGEA in a real world brain atlas, contrasting dissections amongst microglia as less abundant cell type. Metacells were constructed via KNN graphs while tuning hyperparameters. Yield, robustness, and consistency of DGEs together with biological relevance were evaluated.
While cell level analyses maximized sensitivity, they are vulnerable to exteremly inflated false-positive discovery that might not fully be corrected for with multiple testing penalties, even if conservative approaches are used. However, in pharmaceutical research, our goal is not to generate the longest DEG list possible, but pragmatically to generate the most credible list possible, since we aim to identify the genes and biological mechanisms that are most likely to matter in patients.
Real-world data constraints determined statistical feasibility while tuning strongly affected interpretability.
Overall, metacell based DGEA yields reproducible biological signals at competitive cost. MetaDGEA is a promising tool for discovery and assessment of therapeutic concepts to deliver innovation, safety, and quality for our patients.
Co-authors: Martijn van Attekum, Stefano Nardone, Till Andlauer, Stefano Patassini, Maria Faelth-Savitski
Contact Attendee
Warning: Attempt to read property "user_email" on string in /home/1276969.cloudwaysapps.com/ydbgzhdjeq/public_html/wp-content/plugins/my-conference-now/functions.php on line 2826