NOEfinder

Making better use of unused peaks

The NOEfinder program is designed to extract information on intermolecular contacts involving proteins or proteins and other biomolecules that is potentially availale in a collection of unused peaks obtained from a typical 3D 13C-NOESY-HSQC or 15N-NOESY-HSQC NMR experiment. These peaks may contain additional protein assignments but also protein-ligand NOE peaks. The latter are often not a target of structure calculation programs because they focus on resolving the protein structure but they contain valuable information on the orientation of the ligand.

The program attempts to assign these unused peaks by mimicing parts of the automatic assignment routines availabe in popular NMR structure calculation programs using the following execution workflow:

1. Parse information for unused peaks from a Xeasy peaks file (unused column).
2. Parse assignment information from a Sparky project file.
3. Optionally parse the structure of the system under study from a PDB file.
4. Search for carbon-proton (CH) pairs among the unused peaks by by matching the carbon chemical shift with a tolerance of 0.5 ppm and the proton chemical shift with a tolerance of 0.05 ppm in the w2 and w3 dimensions respectively. Tolerance values can be adjusted.
5. For the identified CH couple search for a matching unused proton peak in the w1 dimension with a proton chemical shift tolerance of 0.05 ppm. The identified CHH pair
6. If a structure file was parsed, identify the carbon of the carbon-proton couple identified in step 4 and the proton of the carbon-proton-proton pair identified in step 5. Calculate the distance between the carbon and the latter proton in Angstrom.
7. If the distance is below a predefined cutoff (6A by default) the assignment is labeled as potential intermolecular NOE.
8. Label the CHH pairs as inter- or intra-residue assignment based on the residue numbers. Label the assignment protein-protein (PP) if there are only amino-acids involved or protein-other (PO) if one of the partners is not an amino-acid. The requires a correct atom and residue naming scheme to be used in the assignment file. NOEfinder includes tools to assist on atom and residue translations.
9. If a structure file was parsed and a distance could be calculated for a CHH pair, the information in step 8 can be further specified as through bond (TB), through residue (TR) or inter-residue (IR) assignments.
10. Print peak and assignment information for all identified CHH pairs to the terminal standard output in a table format. This table contains the CHH pair label assigned in step 8 and 9 and the calculated distance if obtained (999 otherwise). The table format can be easily parsed and filtered using standard UNIX command line tools (e.a. grep,awk,sed).

Input:

• peakfile: H-C-H noesy peakfile in Xeasy format
• Sparky project: a Sparky .proj file
• PDB (obtional): PDB structure file

Run: python NOEfinder.py -p -y -s

Output: The NOEfinder program reports results as tabular output written to standard out. The table reports for each assignment the peak ID and the associated carbon and two protons from the w2, w3 and w1 dimensions of the 3D experiment respectively. For each dimension the associated residue name and number, atom name and number and chemical shift are reported. Each assignment is labeled with a four character capitalized string (Cat. column). The first two characters indicate:

• IR: Inter Residue assignment, any C-H-H assignment between two different residues.
• TB: Through Bond, assignment includes atoms covalently linked (calculated distance < 1.09 Å)
• TR: Through Residue, assignment includes atoms part of the same residue (name and number)
• NX: No carbon atom found in structure.

The second two characters indicate:

• PP: Assignment includes protein residues only (standard amino-acids).
• PO: Assignment includes a protein residue and a non-protein residue judged by residue name.
• XX: Not assigned, in case of NC. NC=? = NX?

The label allows for easy sorting and filtering of the output table using standard UNIX command line tools. The last column of the table holds the calculated distances in Å between the carbon atom of the identified carbon-proton pair and the proton in the w1 dimension in case the PDB structure file of the associated protein or complex is provided. If such structure file is not provided or the distance could not be resolved the distance value is set to 999.000

Settings: Basic program settings are defined using the command line interface. Advanced options such as residue and atom renaming conventions (see below) and persistent storage of settings are enabled via the python style configuration file (config_noefinder.py). Please read the documentation on how to use the configuration file in the header of that file.

Residue and atom naming: Identification of residues and atoms in structure files requires their naming to match those of the assignments which is very often not the case. The NOEfinder program offers advanced residue and atom translation services for this purpose. Please consult the config_noefinder.py file for more information on how to use it.

NOEfinder performs this task by:

• Loading unassigned peaks from a 3D NOE experiment (Xeasy format).
• Find carbond-proton assignment pairs for unassigned peaks matching carbon (13C) chemical shift +/- tolerance and proton (1H) chemical shift +/- tolerance in w2 and w3 dimensions respectivly.