Assigning Estradiol - Practical Case

Introduction

NMR spectra are built up from peaks, which are considered as the basic unit of information. A peak in turn is made from transitions. A transition is the observable associated with a jump from one energy level to another in a quantum-mechanical system. A single NMR peak can be constituted by one or many transitions and very often many of them can collapse in one peak and cannot be resolved.

NMR peaks can be grouped into what are known as multiplets: Broadly speaking, a multiplet can be defined as the set of peaks that can be assigned to a particular spin (note: in the case of strongly coupled spin systems, there are the so-called combination lines, which no longer can be assigned to any one nucleus appear). Hence, a multiplet can be just a single transition, a single peak (composed by one or more transitions), or a set of peaks due to the interaction of the assigned spin (or nucleus) with other particles (e.g. due to scalar or dipolar couplings).

From practical reasons, in the context of small molecules, NMR assignments are made from the molecule atoms (e.g. 1H or 13C atoms) to multiplets rather than to individual peaks or even transitions (note: assignments to transitions were popular in the past in NMR spin fitting software packages, such as those following the LAOCOON approach).

In former versions of Mnova, the concept of multiplets existed only for 1D spectra and assignments to 2D spectra had to be made to peaks, integrals or specific spectral regions. Version 14 and higher fill this gap and multiplets can be used for both 1D and 2D spectra. Whilst the other NMR elements (e.g. peaks, integrals) can still be used, it is highly recommended to use always multiplets when assigning 1D and 2D spectra.

The features described below have been described using the this Mnova document

1D assignments

There are several workflows that can be used to assign molecules to 1H (and 13C) NMR spectra, namely:

1)AutoAssignment: Whilst this method usually achieves a >85% of correct assignments, it is usually necessary to correct some potential misassignments. In any case, as a result, protons in the molecule will be assigned to 1D Multiplets in the spectrum

2)Run first Automatic Multiplet Analysis so that the multiplet will be first automatically generated. Next, edit (e.g. delete or add) any multiplet that is not considered necessary or it is missing. Finally, enter into the manual assignments mode (e.g. shortcut: press A or a in the keyboard) and add the assignment.  
 

3)   You can also create the multiplets manually, one by one. Mnova 14 introduces a quick method: once the program enters the manual multiplet mode (e.g. after pressing the J key), as you hover the mouse over the spectrum, the program will highlight the corresponding multiplet. If satisfied with that suggestion, just click and the program will create the multiplet. Of course, you can also manually define the multiplet region following the same procedure used in former versions of Mnova.

Note: In the case of points 3 and 4, the program creates a list of multiplets in the background. Whilst this process is computationally quite efficient, it can take a few seconds when the spectrum has many peaks. As a result, it is possible that when entering into the manual multiplet analysis of assignments, a wait cursor indicating that the program is doing some calculations may show up. Those calculations are carried out in a separate thread so that the Mnova User Interface should still be responsive.

2D assignments

This is probably one of the major novelties introduced in Mnova 14, the concept of 2D Multiplets. This is essentially a direct extension of the idea of 1D multiplet to 2D spectra: all peaks corresponding to the same assignment atom(s) are grouped into one single entity.  For instance, using the same estradiol example as before, this is how the assignment to H4 in a HSQC would look like:

Note that this multiplet is comprised of two peaks which are merged together into one single multiplet. The central coordinates of the multiplet is showed as a pair of dotted cross hairs. Such coordinates are computed by using a weighted average of all peaks within the multiplet.  

The information displayed together with the 2D multiplet box can be customized according to User preferences. For example, the peak position (coordinates), labels, etc, can be turned on/off from the preferences. For example, in the figure below all information about a 2D multiplet is displayed:

If the peaks values and assignments are not needed, they can be changed from the preferences:

With all those options off, the multiplet will look like this:

Finally, the multiplet legend can also be turned off from the preferences:

Resulting in this:

As in the case of 1D multiplets/assignments, the program will automatically highlight a 2D multiplet as the mouse cursor is hovered over the 2D spectrum. Again, at that point, just left click to create the multiplet/assignment.

1D and 2D assignments for the same molecule

Note: It is very important that the 1H and HSQC spectra (in general, ALL spectra) are properly aligned (e.g. referenced) before assigning the spectra.

An important question arises when several spectra are used in the assignment process. For example, let’s consider the simple case of 1H and HSQC. We will have the situation in which, for example, one 1H multiplet is assigned to a particular proton in the molecule and, correspondingly, another multiplet in the HSQC is also assigned to the same 1H spectrum (in the proton dimension). In an ideal situation, the assigned chemical shift for both multiplets will be the same, or very close within experimental error (e.g. digital resolution).

For instance, let’s consider the Estradiol example. If we assign atom 4 in the 1H spectrum, the chemical shift of corresponding 1H multiplet would be 7.024

Next, if we do the same in the HSQC spectrum, the 1H chemical shift of corresponding 2D multiplet will be 7.028 ppm:

So the question is: given those two assigned multiplets, which would be the global assignment chemical shift? Obviously, there can be only one assigned chemical shift that must be computed from the two multiplets. There are a number of potential solutions to this question, but the approach taken in Mnova 14 is the following: always take the chemical shift of the multiplet with the narrowest interval. In the particular case of H4, the difference is really tiny, both multiplets have a very similar interval (HSQC = 0.074 ppm; 1H = 0.066 ppm), but as the 1H multiplet is narrower, the final assigned chemical shift will be taken from the 1H multiplet, that is, 7.024 (from the 1H multiplet) and not 7.028 (from the HSQC multiplet). This is depicted in the figure below:

Note: This behavior can be turned off in the assignments properties so that the assigned value will correspond to the last assigned multiplet. Of course, the table of assigned can be manually edited so that any value can be entered.

The reason why this calculation of the unified chemical shift when several multiplets are available makes sense can be better understood with another couple of multiplets from Estradiol. For example, looking at the aliphatic part of the 1H spectrum, we see one highly crowdy multiplet that can be assigned to several distinct protons, one of them being H9. We can make such assignment:

If we move to the HSQC spectrum, that assignment can be easily located:

Since HSQC multiplet is better resolved than the 1H multiplet and hence, its interval (along the 1H dimension) is shorter, Mnova will take the chemical shift from this multiplet as the final assignment value, as illustrated in the figure below.

Additional new assignment features in Mnova 14

In addition to the features just described, Mnova 14.1 comes with more tools aimed at facilitating the manual assignment of NMR spectra. One of them is that when the Mnova contains a 1H and a HSQC spectra loaded, when one assignment is made (for example, in the 1H spectrum), the program will attempt to create another assignment (and with corresponding multiplet) in the other spectrum (e.g. HSQC) even if no multiplets have been created.

Under the hood, what the program does is calculate a list of multiplets for all spectra in the background so that when a multiplet is needed, the program takes it from there. For instance, assuming that we have a 1H and an HSQC loaded in the program. No multiplets exist in any of the two spectra. We can assign H4 in the 1H spectrum. As soon as this is done, a new 2D multiplet will be created in the HSQC and assigned to the same H4 in the molecule. That is:

1)H4 is assigned in the 1H spectrum:

2)A new assigned 2D multiplet will appear in the HSQC spectrum:

The same workflow can be followed in the other direction: when a 2D multiplet is assigned in the HSQC spectrum, another multiplet will be created in the 1H spectrum and the assignment deduced.