Table 1 PROMO’s main analysis types
From: PROMO: an interactive tool for analyzing clinically-labeled multi-omic cancer datasets
Analysis type | Biomedical goal | Relevant PROMO features | |
|---|---|---|---|
1 | General exploration and visualization | Explore the genomic dataset vis-à-vis the clinical labels Prepare the dataset for downstream analysis, test its consistency and visualize its properties | • Variance-based feature filtering • Label-based sample filtering • Normalization • Sorting by sample label or mean expression • Visualizing data distribution • PCA, t-SNE |
2 | Focus on genes of interest | Explore the expression profiles of specific genes vis-à-vis multiple clinical labels Identify co-expressed genes | • Filter features based on gene symbols • Rank genes by correlation to a given gene symbol • Multi-label matrix visualization |
3 | Disease subtype identification | Look for clinically significant sample clusters | • Sample clustering • Label enrichment analysis • Survival analysis • Classification |
4 | Co-regulated feature group identification | Identify groups of similar features, characterize each group by function | • Feature clustering • GO Enrichment analysis |
5 | Biomarker discovery | Find features that distinguish among sample groups, correlate groups with survival and other clinical data | • Statistical tests for identifying differentially expressed genes • Biomarker-based survival analysis • Rank genes by survival prediction |
6 | Integrative multi-omic analysis | Stratify patients and identify coherent feature groups by integrating data from different omics | • Multi-omic sample clustering • Inter-omic feature correlation |