reportroc





choosefile *  example




1.Task refreshing

2.Parameter adjustment

3.Graphics resizing and downloading

1. function: receiver operator characteristic curve (ROC curve), a curve drawn with true positive rate TPR (sensitivity) as ordinate and false positive rate FPR (specificity) as abscissa. It is mainly used to evaluate the advantages and disadvantages of a binary classifier, and is often used to evaluate the distinguishing effect of a species or a gene between two groups (i.e., indicative analysis of species or genes).

2. Application scope:It is applicable to transcriptome, metabolome, protein, 16S and other data suitable for drawing ROC curve.

3. Input

Data file: Enter two columns of data, the first column is grouping information, and the second column is species abundance or gene expression.



Description of uploaded file format: the format is as follows. The data must be separated by tab characters. You can choose to open it in Excel and save it as a. txt file separated by tab key.


Set task number: enter an identifiable number name; if not, the default number will be used.

4. Output

1) task output

The program calculates and draws according to the input file, and switches tasks in the result display to check the task status.



2) Graphics global parameters:

Graphic title and font can be modified. You can check the display border and diagonal auxiliary line. Scale marks, AUC values, optimal threshold points and coordinates of all drawing points.

3) Color modification

You can adjust the fill color and fill type above and below the curve, and also adjust the thickness and color of the curve.

4) picture download:

It can adjust the height and width of graphics and preview them. Provide SVG and png formats for downloading.

5) Table results

1.factor: enter the column name of the second column of the table

2.AUC/95% CI (AUC): AUC value and its 95% confidence interval range. AUC value is used to measure the discrimination effect of an input species or gene between two groups. Usually: AUC is in the range of 0.5~0.7, and it is considered that the discrimination effect between the model groups is average; AUC is 0.7~0.9, and AUC>0.9, which indicates that this model has a better differentiation effect between groups.

3. best thresholds: the best threshold, that is, when the threshold is set to this value, the classifier has the best distinguishing effect between groups (the best threshold point on the graph is closest to the upper left corner on the curve)



Results Display      (Click " task ID" to view different analysis results)