ORB, A HOMOLOGY-BASED PROGRAM FOR THE PREDICTION OF PROTEIN NMR CHEMICAL-SHIFTS

A computer program (ORB) has been developed to predict H-1, C-13 and N -15 NMR chemical shifts of previously unassigned proteins. The program makes use of the information contained in a chemical shift database o f previously assigned proteins supplemented by a statistically derived averaged chemical shift database in which the shifts are categorized according to their residue, atom and secondary structure type [Wishart et al. (1991) J. Mol. Biol., 222, 311-333]. The prediction process st arts with a multiple alignment of all previously assigned proteins wit h the unassigned query protein. ORB uses the sequence and secondary st ructure alignment program XALIGN for this task [Wishart et al. (1994) CABIOS, 10, 121-132; 687-688]. The prediction algorithm in ORB is base d on a scoring of the known shifts for each sequence. The scores depen d on global sequence similarity, local sequence similarity, structural similarity and residue similarity and determine how much weight one p articular shift is given in the prediction process. In situations wher e no applicable previously assigned chemical shifts are available, the shifts derived from the averaged database are used, In addition to su pplying the user with predicted chemical shifts, ORB calculates a conf idence value for every prediction. These confidence values enable the user to judge which predictions are the most accurate and they are par ticularly useful when ORB is incorporated into a complete autoassignme nt package. The usefulness of ORB was tested on three medium-sized pro teins: an interleukin-8 analog, a troponin C synthetic peptide heterod imer and cardiac troponin C, Excellent results are obtained if ORB is able to use the chemical shifts of al least one highly homologous sequ ence. ORB performs well as long as the sequence identity between prote ins with known chemical shifts and the new sequence is not less than 3 0%.