Problem for quantification of TMT16 experiments

[fstein]

Hi,

using the new philosopher 4.2.1, I have some problems with TMT16. Almost all proteins in the protein.tsv output have a quantification value of 0 for the different TMT channels.
When looking into the psm.tsv output, I see that psm's are quantified as aspected.
Also TMT channels are named as indicated in my annotation.txt file (channel_126 channel_127N channel_127C channel_128N channel_128C channel_129N channel_129C channel_130N channel_130C channel_131N channel_131C channel_132N channel_132C channel_133N channel_133C channel_134N).
However, when looking into the ion.tsv, channels names are wrong and having a "126" as a prefix before the correct channel name (now channel_126126 channel_126127N channel_126127C channel_126128N channel_126128C channel_126129N channel_126129C channel_126130N channel_126130C channel_126131N channel_126131C channel_126132N channel_126132C channel_126133N channel_126133C channel_126134N).
This is also true for peptides.tsv and protein.tsv outputs.
At least for the peptides.tsv and ion.tsv, I have quantified values. So quantification is only missing for protein.tsv output.
Is this information enough to identify a bug?

Or was I doing something wrong?
I am using the pipeline command of philosopher via the command line.

Thanks for looking into this.

Best,

Frank

[prvst]

Can you try running it again but this time without using the annotation file?

[fstein]

Don't you need the annotation file? Or is it just to name the columns?

[prvst]

Thee annotation file is just in case you want to rename the columns to something else besides the channel numbers. You can also rename them manually later if you prefer.

[fstein]

If I leave out the annotation file, I get an error:
Executing label-based quantification on xxx
FATA[...] Cannot read file. open annotation.txt: The system cannot find the file specified.

So apparently, it's not possible to run it without the annotation file :-(

There is definetly, a bug somewhere which was not there in the previous version and also does not appear with TMT11 quantification.

[prvst]

I just notice that you're using the philosopher pipeline in that case, then the annotation is actually needed. Can you share the annotation you're using?

Is there a reason for not using FragPipe? Sorry I don't remember if I ask you this before. The FragPipe workflow will be the go-to option for running the tools we have in the cmd too.

[fstein]

Here is the content of the annotation file:
126 channel_126
127N channel_127N
127C channel_127C
128N channel_128N
128C channel_128C
129N channel_129N
129C channel_129C
130N channel_130N
130C channel_130C
131N channel_131N
131C channel_131C
132N channel_132N
132C channel_132C
133N channel_133N
133C channel_133C
134N channel_134N

[fstein]

I just build our standard pipeline using the pipeline command of philosopher. As I understood, the command line interface for FragPipe is still in the development phase. Therefore, I would like to wait a bit before integrating it.

[fstein]

Did you identify a solution for this already? Sorry for bothering you, but we desperatly depend on TMT16 quantification. If it takes a bit, we need to downgrade again to the previous version in which everything was working (except for TMT6 and TMT18 which we also need desperately).

[fcyu]

Hi @fstein ,

Can you use FragPipe GUI first to make sure that everything is good. Then, you can switch to the headless mode if you want.

ps. The FragPipe headless mode is using the same code base as the FragPipe GUI. There is no difference in terms of the way of running tools and the result files.

Best,

Fengchao

[fstein]

Hi Fengchao,

with the FragPipe GUI it works as expected. I also processed now another TMT16 data set using the pipeline command and there it seems to have worked.
I am now reprocessing the other data set which I mentioned here without adding the variable phospho-modification. This was the only difference to the data set for which it was working just fine.

When do you plan to release the FragPipe headless mode officially?

Best,

Frank

[fcyu]

If FragPipe GUI works, I think the headless mode should also work because they are using the same codebase.

When do you plan to release the FragPipe headless mode officially?

Soon, maybe this month.

Best,

Fengchao

[fstein]

Soon, maybe this month.

That's great. So I will move towards the headless version.

[fstein]

Apparently, TMT16 quantification also did not work without the phosphorylation. I am starting to believe that it's maybe something machine/raw file specific.

What kind of TMT quality filters do you apply for the signal? Do you remove something if it does not fullfill some QC filter?

Do you also remove "bleedthrough" from certain TMT channels to other channels due to chemical impurity of the labels? With the TMT order you get a product data sheet telling you the percentage of impurities which you can use to reduce signal intensity coming from wrong channels.

[anesvi]

You can see the filters in TMT-Integrator tab (Purity score, PSM with low intensity, etc.). We do not apply isotope impurity correction in FragPipe From: fstein ***@***.***> Sent: Thursday, May 5, 2022 12:02 PM To: Nesvilab/philosopher ***@***.***> Cc: Subscribed ***@***.***> Subject: Re: [Nesvilab/philosopher] Problem for quantification of TMT16 experiments (Issue #350) External Email - Use Caution Apparently, TMT16 quantification also did not work without the phosphorylation. I am starting to believe that it's maybe something machine/raw file specific. What kind of TMT quality filters do you apply for the signal? Do you remove something if it does not fullfill some QC filter? Do you also remove "bleedthrough" from certain TMT channels to other channels due to chemical impurity of the labels? With the TMT order you get a product data sheet telling you the percentage of impurities which you can use to reduce signal intensity coming from wrong channels. — Reply to this email directly, view it on GitHub<#350 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AIIMM63IP3X2YP4EO3XIFJDVIPWJJANCNFSM5U3ZOWYQ>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.******@***.***>> ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

[prvst]

@fstein got your email, I'll reply with Alexey in copy. Best

[prvst]

closing this issue because the discussion moved to email.