The Tutorial of TEPITOPEpan Predictions

For the time being, this server online can predict peptides binding to more than 600 other MHC class II moleculars, including the conventional 51 moleculars of Propred(also named TEPITOPE). The steps on how to make a prediction and what is the meaning of the result are as below.

Show an Example Page

A. How to obtain predictions

This is a screenshot of the input page. The numbers in red relate to the referring sections below.

the components of the main frame the main frame press the button to process the prediction
go back

1. Input the protein sequence(s)

In the first place, you should specify the sequence(s) which you want to scan for the binding peptides. The sequences can either be typed directly into the text area, or can be uploaded from a file using the button. If both the text area and the file uploaded containing sequences, this server will treat them both as inputs and the sequence(s) in the uploaded file will be appended to the data in the text area.

Three types of formats are acceptable to this server:

By FASTA Format, long antignic protein sequences in one or multiple lines, each of which begins with '>', are allowed. The lines starting with '>' are treated as the names of the following sequences. In prediction, these sequence are splitted into peptides with the fixed length you specify. The other format, i.e. PEPTIDE, you can input peptides with variable length (at least 9) directly. Then the length option has no sense for the server will give direct predictions for these variable-length peptides.

All sequences have to be amino acids specified in single letter code (ACDEFGHIKLMNPQRSTVWY) and other letter code (BUXZ*) will be pointed out before being ignored by the input processor. The server will ignore any other non-white space characters beside them. If the number of the legal characters in a sequence is under nine, this sequence will be ignored for too short as a binding peptide.

If you have trouble in choosing protein sequences or deciding the suitable format, the web page provides examples for each format. Please click the "show an example" to provide an example for the selected format.

go back

2. Input MHC molecules

The target species can be chose in the first list box. For now only homo species is available.

In the second list box, you can input more than one alleles at a time, the alleles should be space or comma separated. For example,

"DRB1*0101 DRB1*0301 DRB1*0401 DRB5*0205" or "DRB1*0101, DRB1*0301, DRB1*0401, DRB5*0205"

And you can press the button "Check" besides the box to see whether the alleles you input is available at the moment.

In the third box, you can choose the length of the peptides in the protein.

go back

3. Choose output interface

This server provides two types of output interfaces for different users. The "Webpage" interface which is friendly to ordinary users will show all the result on a dynamic web page implemented with javascript and JSP technology. The "Text" interface provides the result directly in a plain text format which is friendly to computer programs.

When all the above have been done, press the "Submit" button to process the prediction.

The time cost of the prediction depends on the number of input data points. The time having passed during the prediction will be shown on the web page.

the state of processing

B. Interpreting prediction output

When the prediction has finished, the processing page will automatically change to the result page.

the discription of the result page the result page click this link to go back
go back

1. Information about prediction

The board shows some basic information during the prediction such as which alleles, the size of the input and how long the prediction processed.

go back

2. The sequences in the input

Each row contains the name and the full sequence of the referring protein.

go back

3. Scores of predictors

The information of prediction scores, percentile ranks and binding cores can be selected and shown by the selecting boxes. Besides, the server can highlight Top scores in "YELLOW" color with different thresholds.

The head of the table is Position, Peptide and the scores of several specific allele predictors. In the position column, the format is "A:B~C" where A means the sequence ID, and B, C mean the start position and the end position of the peptide in the sequence. The second column contains the peptide sequences. The next columns are the scores given by different allele predictors.

The scores are obtained according to specific allele matrix as Propred does. Higher scores means more likely stronger binding between the peptide and allele molecular.

go back

4. Sorting of prediction results

If your browser supports javascript, the Dynamic Table will work. Then you can click the column headers to sort the table and right click the column headers to choose the columns shown on the page.

Sort the table by the scores of SMM-align sort the table
Choose the columns choose the columns of the table