Table 1. List of driver identification methods used to incorporates the prediction of breast cancer driver genes, their working principle and supporting references.

Driver Identification Method	Driver Gene Identification Principle	Citations
IntOGen	Identifies alterations at transcriptomics level, CN gain and losses in tumor sample. It also integrates OncodriveFM for the identification of accumulation mutations, background mutation rate and OncodriveCLUST for mutation cluster identifications. Further, SIFT, Polyphen and Mutation Assessor are used to predict the impact of mutations.	[62–65]
SIFT	Amino acids substitutions and their deleterious impacts prediction. It find the homologous sequences using PSI-BLAST followed by picking sequences with specific diversity and calculating the SIFT scores.	[66]
PolyPhen-2	Analyzes non-synonymous SNP using multiple sequence alignment and structure information followed by predicting the probabilistic damaging variants with confidence prediction and at last interpret the results with mutational impact.	[67]
Mutation Assessor	Predicts mutational impact by calculating functional impact score derived from addition of conservation score and specificity score.	[68]
Driver DBv2	Uses large exome and RNAseq datasets to predict the driver genes using several incorporated tools.	[69]
Active Driver	It identifies significant mutations of cancer genes in active sites of proteins such as mutations in signaling proteins or domains or regulatory elements. It uses gene-centric logistic regression model including multiple factors to estimate mutation significance.	[70]
Dendrix	This algorithm discovers driver genes with high coverage and high specificity using mutation data.	[71, 72]
MDPFinder	It combines mutation and expression data to validate the driver genes and their mutated pathways.	[73]
Simon	It identifies functional mutation impact on proteins, variations in background mutation frequency and genetic code redundancy among tumors.	[74]
NetBox	It identifies the driver genes by comparing genes and performing network analysis on human interaction Network (HIN) data.	[75]
MutSigCV	It uses overall mutation rates and distribution patterns and analyzes background mutation rates with patient specific as well as gene specific mutation rates. Finally it includes expression levels and replication periods.	[76]
MEMo	It identifies the driver genes based on recurrently mutated genes among tumor data with consistent mutational specificity.	[77]
e-Driver	It manipulates internal distribution of somatic functional missense mutations amongst functional domains by relating mutation rates with other regions of same protein.	[78]
DawnRank	Uses gene expression data to construct gene network and rank them based on impact and it analyzes somatic alteration data to identify personalized driver alterations.	[79]
DriverNet	Driver genes are identified based on genomic aberration states of various patients, genes, gene expression data and it further takes biological pathway data into account and builds the network driver genes.	[80]
MSEA	It predicts cancer driver genes based on patterns of mutation hotspot.	[81]
iPAC	Identifies non-random somatic mutations in protein using tertiary protein structure information.	[82]
CoMDP	It uses mutation data to identify driver genes and their pathways. It also predicts genes with other multiple co-occurring biologically significant pathways.	[83]