Identifying spin system types

In ideal situations, a sequential backbone assignment can be made using only sequential connectivities derived from an appropriate set of NMR experiments. However, normally peaks will overlap, some will be missing and multiple assignment possibilities are usually encountered when relying only on sequential connectivities. In addition, mapping stretches of sequentially assigned spin systems onto the primary sequence may be difficult in the absence of peaks, in case of similar residues and repeats of stretches of amino acids in the protein sequence.

To resolve ambiguities and optimise the mapping of stretches of connected spin systems onto the primary sequence, the use of spin system types is often essential. Different amino acids have different side chains that give rise to spin systems varying in the number of connected atoms and chemical shifts. These differences often allow direct identification of an amino acid based on chemical shifts: good examples are:

The deposition of chemical shifts for proteins has allowed for the construction of a database of reference chemical shifts. Comparison of observed chemical shifts and these reference chemical shifts allows prediction of spin system types and secondary structure (e.g. alpha-helix or beta-sheet). In Analysis these reference chemical shifts can be accessed through M: Resonance: Reference Chemical Shifts, which prompts the popup Resonance: Reference Chemical Shifts:

The example shows the CA and side chain carbon chemical shift distributions of Isoleucines.

In order for Analysis to reliably predict the type of residue or spin system type from chemical shifts, it needs to know the atom types corresponding to its chemical shifts. The most indicative shifts that we have obtained so far are those of CA and CB atoms, from the HNCACB and CBCAcoNH spectra. Many automated assignment programs use the CA and CB chemical shifts to obtain residue probabilities for spin systems and map sequentially connected spin systems on the sequence.

Here we illustrate how CA and CB shifts can be used to predict the spin system type manually, and we will use spin system {23}. Use M: Assignment: Pick & Assign From Roots, go to the {Link Peaks} tab, and select spin system {23} (in the # column it is number 106). In the HCN window, show only the HNCACB spectrum:

We know from the information in the CBCAcoNH spectrum, that  two resonances among the HNCACB belong to the sequential residue, and that two resonances among the HNCACB belongs to the CB and the CA of spin system {23}. We will now assign these resonances to atom types: For both the CA and CB peaks of spin system {23} do:

In the HCN window, you should see now that resonances have changed from '[n]' to 'CB[n]', and from '[m]' to 'CA[m]' (CB[669] in the image below):

Investigating spin system {23} in Resonance: Spin Systems shows that it contains 4 resonances with atom types CA, CB, H and N (verify this! M: Resonance: Spin systems). Predicting the type of residue is easy now:

The Spin System Type Scores popup shows up, with a list of residue types sorted and colored according to probability:

In this case it is clear that the most likely type for this spin system is an Alanine. If you are sure about the spin system type, select the row corresponding to the correct spin system type in 'Type Scores' (in this case it is an Alanine) and click [Assign Spin System Type]. The spin system type is added to the spin system information (verify!).

Previous: Assigning non-root resonances

Next: Assigning side chain resonances

Command flowcharts

Looking up reference chemical shifts

Identifying spin system types

Identifying spin system types (alternative way)