Table 2.

PROMO’s key features

Category	Key Features
Data import	▪ Importing genomic data from tabular CSV files ▪ Importing UCSC’s XENA genome matrix and phenotype files ▪ Importing GEO series files ▪ Adding clinical labels from file
Preprocessing	▪ Flooring, ceiling and row normalization ▪ Filtering of samples by clinical labels ▪ Filter features by range, variance, gene symbols or by an external list
Data exploration and visualization	▪ PCA, t-SNE ▪ Data distribution plots ▪ Survival Analysis (Kaplan Meier, Log rank) ▪ Multi-label expression matrix figures
Sorting	▪ Sorting samples and features based on genomic data ▪ Sorting samples based on clinical labels
Clustering	▪ Clustering both samples and features using K-means [27], hierarchical clustering [28], and Click [29] ▪ Browsing clustering history and zooming into specific clusters
Sample cluster analysis	▪ Automated multi-label enrichment test for detecting enrichment of clinical labels
Feature cluster analysis	▪ Gene ontology enrichment analysis
Biomarker discovery	▪ Applying statistical tests for detecting differentially expressed genes/features ▪ Filter results by FDR corrected p-value and fold change ▪ Rank genes based on survival prediction (COX regression)
Classifier generation	▪ Automatic generation of decision tree classifiers for selected sample labels
Integrative multi-omic analysis	▪ Assembly of dataset collection ▪ Multi-omic clustering using SNF [39], NEMO [40] or Consensus Clustering [41] ▪ Inter-omic correlation identification