User Guide
About eHCC-pred (An early HCC diagnosis web server)
This web server was developed to diagnose early HCC with gene expression profiles, which contained mRMR+SVM predictor and MRMD+SVM predictor. The results of MRMD+SVM predictor and mRMR+SVM predictor showed high prediction performance (AUC≥0.92) on both training sets and validation data sets, demonstrating that MRMD+SVM predictor and mRMR+SVM predictor have excellent performance in the diagnosis of early HCC at individual level and they would be helpful for improving clinical decision of HCC diagnosis.
Run Results
Setp1: Upload your file (Select File, .csv)
Setp2: Select feature selection method (mRMR or MRMD)for feature encoding.
New profiles encoded by 1, 0 were generated, where 1 represented Gi > Gj, 0 denoted Gi < Gj, -1 denoted other cases (Gi = Gj ) or ( Gi or Gj does not exist), respectively.
Setp3: Select predictors (MRMD+SVM predictor or mRMR+SVM predictor)
The mRMR+SVM predictor contained 11 gene pairs and MRMD+SVM predictor contained 28 gene pairs.
NOTE:
- (1)For each submission, the number of samples is limited at 100 or less and in csv format.
- (2)For input data, each row represents the expression of a gene, each column represents a sample.
Upload Data
please input .csv file
Note: if there are more than 20 columns, only the first 20 will show here
The order of gene ID has no effect on the prediction results.
Predict
Init the parameter:
Note: if there are more than 20 columns, only the first 20 will show here