MaxQuant v2.7.0 provides unreliable TMT reporter ion intensities from Astral analyser data

[mgraham4]

The TMT reporter ion intensity values from MaxQuant v2.7.0 have many zero values and intensity values that are very different to when the same data is run on a previous version 2.6.3. Although v2.6.3 has no specific support for the Astral, the reporter ion values make more sense and are more typical.

Here are zero values in the reporter ion columns from using v2.7.0 for a TMT 12 plex, run with detection in the Astral analyser, which identified 7,314 protein groups:

Here are the zero values in the reporter ion columns from using v2.6.3 for the same 12 plex, run with detection in the Astral analyser (orbitrap chosen as instrument in MaxQuant), which identified 7,236 protein groups:

The v2.7.0 reporter ion data is also unreliable when no data is missing.
Here are the intensities of the four most intense proteins after using v2.7.0:

Here is the same data using v2.6.3, which makes more sense:

Things I have tried:
Searching data acquired from on a QEPlus with detection in the orbitrap - issue does not occur - issue may be specific to data acquired in the Astral analyser.
Changing to instrument "orbitrap" in v2.7.0 - issue still occurs.

I am using a Windows system.
You should see this same issue if you take TMT data acquired on an Astral analyser and search with v2.7.0.

[JinqiuXiao]

Hi, thanks for letting us know this issue! To help us figure out what happened, can you please send us two mqpar.xml from your Astral run with MaxQuant 2.6.3 and MaxQuant 2.7.0 here? They could only be accessed by us.

Best,
Jinqiu

[mgraham4]

I've uploaded the files. I had to run the 2.6.3 version again because I overwrote the previous file (result was nearly identical).

[JinqiuXiao]

Hi there,

I checked the mqpar files. And found a parameter that might make the difference.

Uncheck the "Use MS1 centroids" in "Group-specific parameters" -> "Instrument" -> "Astral" as the screenshot shows, it was selected in your run.

Also, before you start the run with unchecked "Use MS1 centroids", delete the intermediate files (eg: combined folder) created under the same folder of your raw files, so the new test will not be affected by previous run.

Please let me know if this works for you, thank you!

Best,
Jinqiu

[mgraham4]

Hi Jinqiu
I followed your instructions above. There was only very slight improvement.

I note that "Use MS1 centroids" is the default option for the Astral instrument setting, which is why I used it originally.

[JinqiuXiao]

Hi there,

is it possible to get this dataset for trouble shooting? If it's possible, you can upload your raw files and fasta file here. They could only be accessed by us and used for trouble shooting only.

Best,
Jinqiu

[JinqiuXiao]

Also before you upload the files, have you tried the latest MaxQuant 2.7.3? In MaxQuant 2.7.0 you probably can see for astral the msms.txt file is way bigger than your 2.6.3 run. It's fixed in later versions