Integration-of-Multi-Modal-Omics-Data-for-Breast-Cancer-Patients-Survival-Analysis

Max Rivera and Shreenu Sivakumar

Instructions:

Each of our 3 models (miRNA, mRNA, and miRNA + mRNA) has a main.m file associated with it’s name. For a chosen model, to execute all steps of our workflow, including preprocessing, feature selection, and training, and survival analysis, run the main file.

Main Scripts:

• main_mirna.m = main script for miRNA only model
• main_rnaseq.m = main script for mRNA only model
• main_integrated.m = main script for miRNA + mRNA integrated model

Supporting Functions and Scripts:

• preprocessing_mirna.m = performs preprocessing for miRNA data (no inputs needed)
• preprocessing_rnaseq.m = performs preprocessing for mRNA data (no inputs needed)
• preprocessing_survival.m = performs preprocessing for survival data (no inputs needed)
• featureselection_mirna.m = performs feature selection on preprocessed data (either modality)
• cox_mirna.m = calculates risk scores for patients in testing data (either modality)
• kaplanmeier.m = uses risk scores to split patients into high and low risk groups for visualization
• concordanceIndex.m = uses survival time and risk scores to calculate concordance index
• stepFunction.m = supporting function for concordance index
• tables.m = script that create heat maps and plots to organize performance metric results

Data Files used:

• mirna.tsv = tsv file of miRNA data
• rnaseq-aaa.tsv = tsv file of mRNA
• data survival.tsv = tsv file for survival data