The reason DIA potentially can cover more compounds is because of the duty cycle of the instrument. Each scan cycle takes a certain amount of time, so in that time DDA is inherently limited to whatever N it is set to, and the number of scan cycles that can be achieved within a given compound or set of co-eluting compounds rt. as a result you can identify N compounds per scan cycle, but intense peaks will often be fragmented multiple times in a row, so the practical coverage will be a bit lower than N times the number of scan cycles in a method. with DIA each duty cycle covers the entire m/z range of interest (at the expense of being computationally more challenging to interpret).

Hi Andrew,

Your basic assumptions are first of all correct. However, there are some further details to consider. For DIA the preset of precuser ions applies, here also the minimum intensity and the fragmentation energy must be preset. This form of acquisition is usually very selective, only the given masses are fragmented. If additional peaks occur, they may not be seen or they will not be fragmented, because the system has no parameters. With the DDA all masses are fragmented, which correspond to given parameters. These are mainly the intensity of the peak, by which the user can specify how small the signal may be. The fragmentation energy is then used to generate the fragmentation. Usually a dynamic exclusion is then used. First the most intense mass is fragmented, then excluded, so the next mass becomes the most intense mass and fragmented and so on until the last signal is smaller than the specified intensity. This fragmentation ends with that. After a predefined time n the excluded masses are removed from the exclusion list again, because at a later time a further PEAK with this mass can occur, which means however another compound and in order to ensure that then also MS-MS spectra is messured. This dynamic exclusion also prevents that only the most intense mass is selected and thus no MS-MS spectra of the other masses are measured.

Many greetings

Joachim Horst

The issue was well-described above. DIA is a holistic approach, which means collecting every mass for a given mass range (window) and subjecting them to fragmentation. DDA is a kind of stochastic approach that means for a given mass range, not every mass but only the predefined filtered masses (such as topN=x, top Speed=x, or monoisotopic masses only, charge state between 1-3 only, intensity above....etc) can be subjected to fragmentation. This means many filtered masses cannot be fragmented due to the filter set by the user. Since the DIA divides the total scan range into user-defined scanning windows (pieces), it is inherently more sensitive but more complex data/spectrum is collected. If you have a general idea about diversity and abundance, a matrix or analysis target can give hints about the acquisition selection.

Silas Mellor the senior collegues of my lab for example uses top10. So as I understand, 1 cycle time (say, it takes 1 sec) so then for a portion that 1 second, MS1 is done and for another portion, MS2 of the top10 peaks and therefore the peaks that are "hiding behind" the high peaks miss the opportunity to be fragmented. But I heard there is something called a exclusion time? so for example if there is a 30 second exclusion time, the peaks of the same m/z do not get fragmented. But my question is, after the 30 sec exclusion, the peak that was "missed" will too eventually get fragmented so then no stone will be left unturned no? just like the DIA's goal is to leave no stone unturned?

I think I'm not grasping something important idk what it is....

Joachim Horst but eventually after a set dynamic exclusion time, we will eventually get to the msms spectra of the "other masses" no? Say there's a peak at m/z 2000 and it gets MS2-ed, then for a set amount of time, the mass spectrometer is going to assume any other peaks that appear at 2000 is redundant, but after that, wouldn't the mass spectrometer eventually be like "oh shoot, I missed a peak!" and msms the lesser peaks that appear in m/z 2000?

excuse my somewhat anthropomorphic language. Mass spectrometry is still new to me lol

The thing is with UPLC, which is commonly used with MS instruments nowadays, chromatographic peak widths tend to be a lot less than 30s, sp the peaks that are missed by the instrument will be come and gone by the time it gets to them. Dynamic exclusion helps a bit to avoid the instrument just sampling the same compound again and again, but you still miss things just because of the inherently stochastic nature of the process and the fact that you have a finite amount of time to measure a given peak.

Hello Andrew,

a mass spectrometer always does what is described in the method.

There it is specified how many spectra / second are recorded, how often the MS-MS Spetrum is repeated and how long the whole cycle should last.

For example, if the mass 2000 exceeds the minimum intensity, the MS-MS spectrum is generated. The method then specifies how often the MS/ MS spectrum is repeated, usually 2- 3 times. Then the exclusion is done for N seconds.

Modern mass spectrometers have acquisition rates of 10-50 spectra / second, so that with a LC peak of 20 seconds there is enough time to acquire further MS-MS spectra.

I then make the exclusion time dependent on how long my LC peak will last, in my case rarely longer than 20 seconds. If, as in your example, the mass 2000 still exceeds the minimum intensity after this time, the MS-MS spectrum is generated again. Essentially, the minimum intensity is important in the setup, since I do not want to fragment the background masses, but only the chromatographic peaks. The cycle times for all spectra should also not be too large, to give the MS enough time for the spectra and to get enough data points to plot the chromatogram. 1-3 seconds is common for us.

Since a spectrum lasts a few milliseconds, there is enough time for the MS-MS experiments so that normally no mass above the minimum intensity is "missed". The top10 method of your colleagues we usually call Auto-MS-MS.

A typical setup for MS^3 Methods at our facility would be for our Orbitrap for example:

Fullscan spectrum 150- 1500u

peak intensity > 10^6 --> 2 x MS-MS experiment with 30 eV

(MS^2)-->recording spectrum

exclusion of mass for 20 sec

-->peak intesity > 10^5 -->2 x MS-MS experiment of the product ions

with 45 eV

(MS^3)-->recording of spectra

exclusion of masses for 15 sec

cycle time 2 sec for all experiments

So for a 20 sec LC peak we generate about 10 fullscan spectra with the corresponding number of MS^3 spectra.

I hope I was able to reduce your concerns about masses being "overlooked" a little.

Many greetings

Joachim

Joachim Horst ohhh now I'm getting the picture (albeit, a little) so its sort of like balancing between the minimum intensity and the background peaks; you don't want background peaks but naturally, some low abundant peaks will not exceed the intensity and therefore will be missed, so that's where DIA comes in, where EVERYTHING is fragmented but you need a spectral library to sort which peaks come from what precursor..?

Thanks! I think I'm slowly starting to understand!