Membrane Protein Secondary Structure Prediction Server

Transmembrane, transmembrane structure, protein transmembrane structure, transmembrane structure prediction, protein transmembrane structure prediction, membrane protein structure, structure prediction, protein structure prediction, protein secondary structure, protein secondary structure prediction, membrane protein structure prediction, membrane protein secondary structure prediction, preference functions, hydrophobicity analysis, sequence, protein sequence, amino acid scales, hydrophobicity scales, hydrophobicity plot, biochemistry, biophysics, biocomputing, scientific computational server, Davor Juretic, Damir Zucic, Ana Jeroncic.
The purpose of this server is to predict the transmembrane (TM) secondary structures of membrane proteins, using the method of preference functions. Scientific publications describing the method are listed at the bottom of this page. The method was invented by Davor Juretic , professor at the University of Split , Croatia. This server was written by Damir Zucic, at the University of Osijek , Croatia. See the page bottom for technical details about the server. Ana Jeroncic was involved both in development of the prediction program and in testing of this server.

Terms of Use

(1) Academic, educational and any other noncommercial use of this software is free of charge.
(2) Commercial use of this software has to be negotiated with the project leader, prof. dr. Davor Juretic (see the address at the page bottom), except for the trial period of two days. The trial period is counted per site, not per user.
(3) Every access to this server is logged. If you don't like this, please disconnect now.
(4) The authors retain the right to change the licensing policy. Be sure to read the latest terms of use!

Technical Requirements

Your browser may be connected to the Internet through a proxy server. If this is the case, you will be unable to retrieve the correct prediction results. Change the configuration of your browser to avoid this problem. This should be very easy; if you don't know how to switch off the proxy server, ask your local administrator or any person familiar with browser configuration.

Data Input Form

Please paste or type your sequence (one letter code) in the frame below. Do not submit the nucleotide sequence! You can submit the sequence in a free, case independent format, just don't forget to remove any extra information (like protein name, code etc).

There are three ways to copy and paste the sequence:

(1) Use an editor to read the sequence file, stored locally on your computer, and copy data to browser window.

(2) Use the SWISS-PROT search to retrieve the sequence. You will need an additional browser window , connected to the SWISS-PROT search engine. If you have no intention to combine SPLIT server with SWISS-PROT right now, try it some other time, you will like it.

(3) If you have no sequence and you don't want to search the SWISS-PROT database, use sample sequences . If you just want to test the SPLIT server this is the easiest way to prepare data.

Membrane protein sequence is the minimal input required by the prediction program.

Type the name of your protein in the small frame below. This input is not required for prediction, so you can ignore it.

Your personal or institutions home page; not required, but we would be pleased if you invite us to visit your pages.

If you specified the sequence, you can press the submit button to start the prediction. If you want to specify some extra information, fill the rest of this form. You may experiment with these extra parameters, but if you haven't read the references it is better to use default values.

Program Version

Two versions of the structure prediction program are available: split35 and split40. Versions split31 and split35 are described in the publications listed at the bottom of this page. Default version is split35. If you want to use newer (split 4.0) wersion, please click here

Amino Acid Attribute Scales

Here you can choose amino acid scales to be used in calculation. Click here for description of amino acid scales.
If you are interested in numeric values, two pages are available: A compact one (12 kB), and a big one (106 kB).
The second is easier to read, but if your connection to our site is slow, load the compact page. Both pages contain the same numeric data.

Amino acid scales
for calculation of


Amino acid scales
for calculation of


Additional Input For Version 35

An extra scale for calculating conformational preferences (number 60) can be used to extend N and C terminus of predicted transmembrane helices.

Use extra scale No extra scale

Now you can submit your data. Press SUBMIT button to start the prediction. This submit button has exactly the same function as the similar button above (the same applies to clear button).

Technical Notes

The prediction program was written in FORTRAN 77. The rest of the server was written in HTML, version 3 (this form, for example, and all other HTML documents), ANSI C (output data conversion and graphics) and unix shell script language (CGI script).

Most of development was done on personal computers running Linux (more accurately called GNU /Linux) operating system. FORTRAN to C translator (D. Gay, S. Feldman, M. Maimone and N. Schryer, copyright AT&T Bell Laboratories and Bellcore) was used to translate the FORTRAN code. C sources were compiled using GNU C compiler (Free Software Foundation) , version Almost all parts of this server are portable across different unix and unix compatible platforms. The graphics library can be used to plot other types of scientific data.

This instance of the server is installed on GNU/Linux computer splitbioinf.pmfst.hr.

Contact us!

We have long-term plans to further develop both our prediction method and this server. Any feedback about your experience with SPLIT server will be of great value for us. We are open for your criticism, suggestions, comments and questions. Feel free to contact us at:

juretic@mapmf.pmfst.hr zucic@garlic.mefos.hr
Davor Juretic
Fakultet PMZOP
Teslina 12
21000 Split
Damir Zucic
Elektrotehnicki fakultet
Kneza Trpimira 2B
31000 Osijek

Related Web Servers

We have a list of other web servers for membrane protein secondary structure prediction. If you are maintaing the similar server, and are willing to be added to our list, please notify us. Send the short description, with precise address of your server, to Damir Zucic (zucic@garlic.mefos.hr) .


This server was partly financed by the Croatian Ministry of Science and Technology. Igor Susenka and Ivica Loncar, system administrators at the Faculty of Electrotechnics, University of Osijek, provided valuable help during development of this server.


Older reference:

Juretic, D., Lee, B. K., Trinajstic, N., and Williams, R. W. "Conformational preference functions for predicting helices in membrane proteins."
Biopolymers 33, 255-273 (1993).

Recent references:

1) Juretic, D., Lucic, B., Zucic, D. and Trinajstic, N. "Protein transmembrane structure: recognition and prediction by using hydrophobicity scales through preference functions."
Theoretical and Computational Chemistry, Vol.5. Theoretical Organic Chemistry, pp. 405-445. Editor: Parkanyi, C., Elsevier Science, Amsterdam, 1998.
SPLIT 3.1 results are described in that paper.

2) Juretic, D., Zucic, D., Lucic, B. and Trinajstic, N. "Preference functions for prediction of membrane-buried helices in integral membrane proteins."
Computers Chem. Vol. 22, No. 4, pp. 279-294, 1998.
SPLIT 3.1 results are described in that paper as well.

3) Juretic, D. and Lucin A. "The preference functions method for predicting protein helical turns with membrane propensity."
J. Chem. Inf. Comput. Sci. Vol. 38, No. 4, pp. 575-585, 1998.
SPLIT 3.5 results are described in that paper.

4) Juretic, D., Jeroncic, A. and Zucic, D. "Sequence analysis of membrane proteins with the web server SPLIT." Croatica Chemica Acta Vol. 72, No. 4, pp. 975-997, 1999.

5) Juretic, D., Jeroncic, A. and Zucic, D. "Prediction of initiation sites for protein folding." Periodicum Biologorum 101, No 4, 339-347, 1999.