DeepDoublet: A Tool for Analyzing Cell Interactions
Author Information
Author(s): Liao Linbu, Kim Junyoung, Cho Kanghee, Kim Junil, Lim Byung-Kwan, Won Kyoung Jae
Primary Institution: Cedars-Sinai Medical Center
Hypothesis
Cell interaction can change the transcriptome of a cell.
Conclusion
DeepDoublet successfully identifies neighboring cell-dependent gene expression, particularly in hepatocytes interacting with liver endothelial cells.
Supporting Evidence
- DeepDoublet identified specific expression of hepatocytes when interacting with liver endothelial cells.
- Angptl3 was found to be highly expressed in hepatocytes interacting with LECs.
- The tool was trained using 400,000 artificial doublets generated from hepatocytes and LECs.
Takeaway
DeepDoublet is like a smart tool that helps scientists understand how two types of cells talk to each other and how that affects their behavior.
Methodology
DeepDoublet uses deep learning to analyze transcriptomes of physically interacting cells.
Potential Biases
Potential bias in the selection of cell types and the artificial doublets used for training.
Limitations
The study primarily focused on hepatocytes and LECs, limiting the generalizability to other cell types.
Participant Demographics
Adult mouse liver cells were used in the study.
Statistical Information
P-Value
0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website