CAMTP Center for Applied Mathematics and Theoretical Physics University of Maribor, Slovenia

Protein Structure Prediction

Determination of 3D protein structure is of great importance for prediction of their function in farmacology and in medicine. So far, there are over 50.000 known proteins, but the structure is known for around 500 only. Experimental methods for protein structure determination (X-ray diffraction, NMR) are expensive and time consuming.

We present a set of available databases and software tools which help in protein structure prediction. An emphasis is on the use of Inductive Logic Programming (ILP, machine learning) for the secondary structure prediction. ILP is applied to learn rules from the protein data base of known structures. Additionaly, we show how Constraint Logic Programming is combined with classical methods of macromolecular simulation to determine the minimum energy structure of protein molecules.

We took part in the "Third Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction", CASP3, at Asilomar, USA, 13-17 December 1998. The results of the secondary protein structure predictions are available on-line.

The following slides were presented at the "Biomedical Applications of Computational Logic and Machine Learning" advanced seminar, Bled, 11 November 1997:

slide00	slide01	slide02	slide03
slide04	slide05	slide06	slide07
slide08	slide09	slide10	slide11
slide11a	slide12	slide13	slide14
slide15	slide16	slide17	slide18
slide18a	slide19	slide19a	slide20
slide21	slide22	slide23	slide24
slide25	slide26	slide27	slide28
slide29

Last modified: 31 Dec 1998 by Igor Mozetic