LACS

From NESG Wiki
Revision as of 17:15, 20 November 2009 by Prossi (talk | contribs)
Jump to navigation Jump to search

Linear analysis of chemical shift (LACS) provides an independent means for absolute referencing of protein chemical shifts.  LACS is routinely run on all chemical shift sets submitted to the BioMagResBank (BMRB)   In the contest of protein structure determination chemical shifts (CS) play a key role as reporters of secondary structure.  Backbone chemical shift derived dihedral angles can be obtained through programs such as TALOS+ and are used as constraints in the structure determination process. The backbone shift set (CA, CB, CO, HN, N, HA) is also used in the molecular fragment selection (MFR) for CS-Rosetta structure prediction protocol.  Here fragments are selected from the protein databank (PDB) that have the same predicted CS as the query sequence.  Therefore, successful use of chemical shift for secondary structure inference requires the correct referencing.  LACS server gives independent CS offsets for CA, CB, CO and HA and allows for effective correction of chemical shift sets derived from [U-13C,15N] and from [U-2H,13C,15N].  The latter labeling scheme may show distinct degrees of offset for CA and CB due to the number of bound deuterons.   

LACS can be run on the native server using a chemical shift file in bmrb 2.1 format and the user's email and project name.  The server returns the results within a few minutes when the correct chemical shift format are provided.  The chemical shift file format is stringent and the output is of straight forward interpretation.  A running bmrb file and output example are provided for user's convenience, adhering to the format should provide smooth operation of the server.




In the NESG the IUPAC referencing is obtained using internal 50 uM DSS and setting the DSS methyl peak to 0.00 ppm at the temperature of interest.  The spectrometer temperature is calibrated using a neat methanol sample in the 10-30 deg C range.  Under these conditions, backbone CS of double labeled protein samples derived from triple resonance experiments alone typically show small deviations 0.1 - 0.3 ppm in 13C and 0.01 - 0.03 ppm in 1H.  The final CS set derived from double labeled samples filtered through higher resolution [1H,13C]-HSQC tends to give offsets at the lower extreme of this range.  On a smaller statistical sampling of triple labeled samples we found higher offset values (0.4 - 0.8 ppm) in CA and CB.  Offsett correction prior to TALOS or MFR runs are therefore highly recommended especially fro treatment of triple labelled samples.