Compare PLIERbase vs PLIERfull: Alignment of Prior Knowledge (ARI) in GTEx or ARCHS4

[msubirana]

This issue proposes an analysis to compare the alignment of prior knowledge (using Adjusted Rand Index, ARI) between PLIERbase and PLIERfull models, focusing on datasets such as GTEx or ARCHS4.

Objectives:

• Quantitatively compare the alignment of prior knowledge (ARI) between PLIERbase and PLIERfull.
• Perform the comparison on either GTEx or ARCHS4 datasets.
• Summarize findings and provide recommendations for the choice between PLIERbase and PLIERfull in similar analyses.

Tasks:

1. Run both PLIERbase and PLIERfull on selected dataset (GTEx or ARCHS4).
2. Calculate ARI for alignment with prior knowledge for each model.
3. Visualize and compare the results.
4. Document methodology and results in the repository.

Expected Outcome:

• Clear quantitative and visual comparison of ARI for both models.
• Guidance for users on model selection based on prior knowledge alignment.