Fully automated structure determinations of the Fes SH2 domain using different sets of NMR spectra

Abstract

The recently introduced fully automated protein NMR structure determination algorithm (FLYA) yields, without any human intervention, a three-dimensional (3D) protein structure starting from a set of two- and three-dimensional NMR spectra. This paper investigates the influence of reduced sets of experimental spectra on the quality of NMR structures obtained with FLYA. In a case study using the Src homology domain 2 from the human feline sarcoma oncogene Fes (Fes SH2), five reduced data sets selected from the full set of 13 three-dimensional spectra of the previously determined conventional structure were used to calculate the protein structure. Three reduced data sets utilized only CBCA(CO)NH and CBCANH for the backbone assignments and two data sets used only CBCA(CO)NH. All, some, or none of the five original side-chain assignment spectra were used. Results were compared with those of a FLYA calculation for the complete set of spectra and those of the conventionally determined structure. In four of the five cases tested, the three-dimensional structures deviated by less than 1.3 A in backbone RMSD from the conventionally determined Fes SH2 reference structure, showing that the FLYA algorithm is remarkably stable and accurate when used with reduced sets of input spectra.

Evaluation of Multi-Objective Optimization Algorithms for NMR Chemical Shift Assignment

“…We performed experiments with the NMR dataset of a protein for which an NMR structure determination was completed earlier by conventional methods: the 114-residue Src homology 2 domain from the human feline sarcoma oncogene Fes (Fes SH2) [ 43 , 44 , 45 ]. We divided this protein into stretches of 30 residues in order to evaluate the behavior of our algorithms in different datasets.…”

“… 1 Assignment calculations were performed for 7 fragments ( Table 1 ) of Fes SH2 [ 43 , 44 , 45 ]. The total number of correct, wrong and unassigned atoms are given.…”

Structure of the putative 32 kDa myrosinase‐binding protein from Arabidopsis (At3g16450.1) determined by SAIL‐NMR

“…The spectral simplification achieved by SAIL with this larger protein makes it possible to use semi‐ or fully automated methods developed for use with smaller proteins to analyze the NMR data. We are developing a software package that exploits the benefits of the SAIL method [25–27]. Finally, the SAIL method is expected to enable functional investigations of larger proteins.…”

Automated protein structure calculation from NMR data

“…A collection of routines (eg CALIBA, GLOMSA, HABAS) were combined together with a torsional angle molecular dynamics routine to make the program DIANA, which then made use of further routines (REDAC, ASNO, the assignment program GARANT, and molecular dynamics program OPAL) and evolved into DYANA ), which in turn absorbed further routines, in particular the NOE assignment module CANDID, and was packaged within an updated torsion angle molecular dynamics program to become the widely used structure calculation program CYANA (Güntert 2003 FLYA is the only package that claims, with some justification (Scott et al 2006), to be able to calculate protein structures from NMR spectra (but not raw time-domain data) in a fully automatic way. It is based on its demonstrably successful predecessors DYANA and CANDID, as described above, and has so far only been tested on three small (< 16 kDa) proteins (López-Méndez and Güntert 2006).…”