Single-cell-analysis

The analysed file is an RDS object, a list containing:

• Count: A matrix of single cells against genes.
• Metadata: Metadata corresponding to the count matrix, including cell type, label, and replicate.

This dataset has 2 conditions (Disease vs Control), 15 cell types, and 5 replicates.

Task 1

Evaluate biological insights from single-cell data with perturbations:

1. Identify differentially expressed genes between the 2 conditions. Provide a final table with:
   • Cell type
   • Gene
   • Effect size (log-2 fold change)
   • P-value
   • Adjusted p-value
2. Rank cell types by how much they are perturbed, from most to least.

You may use any methods or propose your own for these analyses. Additional insights are welcome.

Task 2

From a translational perspective, explore single-cell sequencing for phenotypic classification:

1. Build a classifier to predict disease vs control. Consider:

   • Should the classifier be cell-type specific?
   • Train-test split: replicate-based or cell-based?

2. Provide results for your best classifier:

   • Accuracy
   • Area under the ROC
   • Confusion matrix

For both tasks, please include scripts and relevant data visualizations.