Skip to main content

Table 4 Some genes and their dysfunctions from the interactions of selected feature subsets of different methods

From: Robust biomarker discovery for hepatocellular carcinoma from high-throughput data by multiple feature selection methods

Gene

Subset

Function

SKAP1

6 methods

SKAP1 encodes a T cell adaptor protein and it is involved in HCC signaling pathways [32, 33]

EPHB1

SVM, KNN, NN

Ephrin-B1 participates in the tumor progression through promoting the formation of new vessels of HCC [36]

STC2

NN, SVM, KNN

STC2 is overexpressed in HCC and acts as a potential oncoprotein [37]

CDHR2

NN, SVM, KNN

CDHR2 is highly expressed in HCC para-carcinoma tissue, but is weakly expressed in tumors. It is found to inhibit tumor growth [38]

FAM134B

NN, SVM, KNN

FAM134B works as a tumor inhibitor and inhibits cancer growth in vitro and in vivo [39]

MUC6

RF, NN, SVM, KNN

MUC6 encodes a member of the mucin protein family. It is a biomarker gene of many cancers [40]

PHOSPHO1

Adaboost, NB, NN, SVM, KNN, Stepwise Regression

PHOSPHO1 is associated with hepatitis B [41]

OXT

NN, SVM, KNN, Stepwise Regression

OXT is found to regulate cell proliferation. It is a key differential gene in nonalcoholic fatty liver disease [42]