Graphical Dirichlet Process for Clustering Non-Exchangeable Grouped Data
Author Information
Author(s): Arhit Chakrabarti, Ni Yang, Ellen Ruth A. Morris, Michael L. Salinas, Robert S. Chapkin, Bani K. Mallick
Primary Institution: Texas A&M University
Hypothesis
Can a Bayesian nonparametric approach effectively cluster non-exchangeable grouped data using a directed acyclic graph?
Conclusion
The proposed graphical Dirichlet process successfully clusters cells based on gene expression while allowing information sharing across non-exchangeable groups.
Supporting Evidence
- The model allows for the sharing of clusters among non-exchangeable groups.
- Simulations demonstrated the model's effectiveness in identifying clusters.
- The method was applied to real single-cell RNA-sequencing data.
Takeaway
This study introduces a new way to group data that doesn't follow the usual rules, helping scientists better understand complex biological systems.
Methodology
The study employs a Bayesian nonparametric model to cluster grouped data characterized by dependencies represented by a directed acyclic graph.
Limitations
The model's performance may vary with different DAG structures and assumptions about the underlying distributions.
Participant Demographics
The study involved mice with varying genetic backgrounds and dietary conditions.
Want to read the original?
Access the complete publication on the publisher's website