CRC Radiotherapy

This repository contains the code used for the project. The project involves data preprocessing and model building, including MOFA (Multi-Omics Factor Analysis) and Random Forest model.

---

Overview of Scripts

The scripts are organised to follow the workflow of the project:

1. Data Preparation

• data_split.Rmd
  Splits the dataset into two subsets:
  • One subset for training the MOFA model.
  • One subset for training the Random Forest model.

2. Data Preprocessing

• RNA_preprocessing.Rmd
  Preprocessing of the RNA dataset.
• mutation_preprocessing.Rmd
  Preprocessing of the mutational dataset.
• methylation_preprocessing.Rmd
  Preprocessing of the methylation dataset.
• cna_preprocessing.Rmd
  Preprocessing of the Copy Number Alteration (CNA) dataset.

3. MOFA Model Development

• MOFA_models.Rmd
  • Identify the optimal MOFA model.
  • Build the optimal MOFA model.
  • Characterize the factors in the model.

4. Feature Extraction

• extract_features.Rmd
  Extract informative features from the MOFA model, particularly in relation to the treatment response covariate.

5. Functional Analysis

• GSEA.Rmd
  Perform Gene Set Enrichment Analysis (GSEA) using important factors identified from the MOFA model.
• pathway_enrichment.Rmd
  Conduct pathway enrichment analysis using the informative factors identified from the MOFA model.

6. Machine Learning

• Random_Forest.ipynb
  • Build and evaluate a Random Forest model using the features extracted from the MOFA model.

Data avalibility

• Preprocessed datasets are provided here to allow running of the later scripts as well as a construced MOFA model.

Preprocessed Datasets

1. RNA Datasets

• RNA_supervised_preprocessed
• RNA_unsupervised_preprocessed

2. CNA Datasets

• CNA_supervised_preprocessed
• CNA_unsupervised_preprocessed

3. Mutation Datasets

• mutation_supervised_preprocessed
• mutation_unsupervised_preprocessed

4. Methylation Datasets

• methylation_supervised_preprocessed
• methylation_unsupervised_preprocessed

---