I have no experience with AB 4600 but I am very satisfied with QExactive. The main advantages of the orbitrap are (in my point of view) the high mass accuracy and the possibility to scan with ultra high resolution.

Usually I have a mass accuracy better than 0,2 ppm. It is very helpfull when you want to perfom database searches in e.g. Mascot and identify proteins after tryptic digestions.

The calibration and handling of the thermo instrument is very easy. Calibration is done automatically and suprisingly the automatic calibration and tuning process works. So I can wholeheartly suggest the QExactive as versatile high performance mass spec.

The calibration is *supposed* to be daily on the QExactive, but we didn't calibrate it until we gave it a deep clean after about a month of running and we had no problems.

I don't know about the 4600 - we've just had a 5600 installed, and a 5600+ is on the way. But we're very happy with the QExactive so far (before we were running a lot of samples on the Orbitrap Classic, would you believe). The QExactive is easier to run but conversely more difficult to alter settings with; however we haven't seen a need to alter anything other than the methods Thermo have described for us.

I'm in the process of gathering information for an instrument purchase, and we're looking at the 5600 and the QExactive. I'd love to get some detailed information about the number of protein IDs between the two instruments and what a typical duty cycle looks like. I have some experience with a 5600, and I know it typically gets around 500 more protein IDs than an Orbitrap Elite (intel from my former boss who has both). How does the QExactive compare?

We haven't got our 5600 in full working mode yet, we did have someone who actually got a higher number of protein IDs out than the QExactive with a test sample, about 500 sounds right.

In terms of the duty cycle, the calibration of the machines is very different. The QExactive should be calibrated every 2 days but you can get away with doing it once a week - however the 5600, we have been told, needs to be calibrated every 1-2 SAMPLES and it seems the mass drift is actually quite severe actually a couple of runs, in the tens of ppm. So for the way our lab operates - generally I do 24-well Offgel fractionation of my samples - this is quite a serious increase in machine running time.

More than that I can't say at the moment, it may be a matter of weeks before I'm able to correct samples directly.

Thank you very much for the information. I will be most interested to hear follow-up on the stability of the calibration. That's something I've had in the back of my mind as a concern. I've had experience with a Q-Tof from a dfferent vendor, and its calibration would also drift badly throughout the day. If the Q Exactive will produce similar results with more stable calibration, then that's a serious consideration.

Following up: by duty cycle, I was referring to the number of MS and MS/MS scans in one second. From other sources I'm hearing 10 as the number for the Q Exactive, and the number is around 50 for the 5600. That accounts for the difference in protein IDs.

I have a QExactive that has been running for 3 months now, and we only do the calibration every two weeks. I always evaluate the mass calibration before doing the recalibration just so I can evaluate the drift, and the worse I have seen is 0.7 ppm. So really, even doing a calibration every two weeks appears to be overkill. The system is very stable.

A colleague of mine bought the 5600, and we compared the same whole cell digest on both instrument on a 100min gradient and the results were very similar (around 1200 IDs).

If the 5600 really has to be recalibrated every few samples, that would be a major drawback. You should get this confirmed by someone that has been running the instrument for a while.

There is a recent paper, that compares sample preparation methods for label-free proteomics. It is anyways an interesting read, and on a side-note they also touch upon your issue: http://f1000research.com/articles/2-272/v2

In figure 2d inlay, they show that they have 30% more identifications of yeast peptides when using DDA on a Qexactive than on a AB5600 tripleToF. We have made similar observations using plant samples at our own machines.

I wonder myself what causes this, as the AB5600 acquires 50 spectra per second. The Qexactive should have only 12hz (I stil don't understand how anyone could do DIA SWATH with this machine...). It is a very intersting question whether the calibration routine is the cause for this difference.

EDIT: ah, this question is a year old. anyways, that paper is still an intersting read. maybe it is valuable to you.

Nice to see such good discussion. I am planning to buy a similar machine but for small molecules, instead of protein samples. I have two questions: 1, why no one mentioned Waters products, such as Xevo G2-S QTOF? 2, how about Thermo Exact plus without Quard modular? My budget is good for these kind of models.

Thanks,