Pseudotime Estimation and ML Pipeline

Overview

This R script encapsulates a pipeline for estimating pseudotime using Seurat and Monocle, followed by training a machine learning model (XGBoost) to predict pseudotime and identifying important genes associated with the process.

Dependencies

Make sure you have the following R packages installed:

• Seurat
• monocle3
• caret
• xgboost

Usage

1. Clone the repository:

   git clone https://github.com/mode1990/Pseudotime-downstream-by-ML
   cd Pseudotime-downstream-by-ML
   
   

Example usage:

Load required libraries

library(Seurat) library(monocle3) library(caret) library(xgboost)

Run the pipeline with your Seurat object path

results <- run_pseudotime_pipeline("/path/to/your/seurat_object.rds")

Access results

print(results$xgb_rmse) print(results$r2) print(results$top_genes) print(results$emb_time_terms) print(results$gene_fits)