Importance of NMR data validation prior to structure calculation

M. Zachrdla, L. Žídek, V. Papoušková, P. Padrta, V. Sklenář

NCBR, Masaryk University, Kamenice 5, 62500 Brno

324489@mail.muni.cz

Quality of structures calculated using NMR restraints always reflects the quality of the experimental data. In this poster, impact of the experimantal error of residual dipolar couplings (RDCs) on calculated structures was analysed by evaluating several parameters describing structure quality and the course of the calculation. The analysis showed that the use of RDCs containing experimental errors results in considerable deterioration of evaluated parameters instead of improving calculated structures compared to calculations without using of RDCs. RDCs, from which the evidently incorrectly measured data were removed, do not deteriorate calculated structures so considerably and in the case of RMSD, better results are achieved than without RDCs, but their impact is still more or less negative. Only RDCs that passed the tests developed in our laboratory, i.e., data with a minimal presence of experimantal errors, provide in some cases better results of evaluated parameters then calculations without the use of RDCs. Our results document that validation of measured RDCs before calculating protein structure is necessary in order to obtain structures of the highest possible quality.