proteinSummarization command error

[Alison Ross]

Hello, I get the following error using the proteinSummarization command:

"Error in vecseq(f__, len__, if (allow.cartesian || notjoin || !anyDuplicated(f__,  : 

  Join results in 4616988 rows; more than 3620460 = nrow(x)+nrow(i). Check for duplicate key values in i each of which join to the same group in x over and over again. If that's ok, try by=.EACHI to run j for each group to avoid the large allocation. If you are sure you wish to proceed, rerun with allow.cartesian=TRUE. Otherwise, please search for this error message in the FAQ, Wiki, Stack Overflow and data.table issue tracker for advice."

evidence: https://drive.google.com/file/d/1WbJUB7QSrAj1g7EnRYVt2nFgDf4RRRD8/view?usp=sharing

[Alison Ross]

proteinGroups: https://drive.google.com/file/d/1ldkp5JNcWo1lv-m8NckJpaHUFcqmMRCB/view?usp=sharing

[Alison Ross]

It's actually an issue with the MQtoMSStatsTMT command

input.mq <- MaxQtoMSstatsTMTFormat(evidence, proteinGroups, annotation.mq)

leads to 

Error in vecseq(f__, len__, if (allow.cartesian || notjoin || !anyDuplicated(f__,  : 

  Join results in 4533939 rows; more than 3620459 = nrow(x)+nrow(i). Check for duplicate key values in i each of which join to the same group in x over and over again. If that's ok, try by=.EACHI to run j for each group to avoid the large allocation. If you are sure you wish to proceed, rerun with allow.cartesian=TRUE. Otherwise, please search for this error message in the FAQ, Wiki, Stack Overflow and data.table issue tracker for advice.

On Monday, August 30, 2021 at 11:54:30 AM UTC-4 Alison Ross wrote:

proteinGroups: https://drive.google.com/file/d/1ldkp5JNcWo1lv-m8NckJpaHUFcqmMRCB/view?usp=sharing

[thuan...@gmail.com]

What version of MSstatsTMT and MSstatsConvert do you use? If you install MSstatsTMT 2.0.1 and MSstatsConvert 1.2.2, the error should be fixed.

-Ting

[Alison Ross]

Hi Ting, I am using MSstatsConvert_1.2.2 and MSstatsTMT_2.0.1, but the error is still persisting :( 

[Mateusz Staniak]

Hi,

let's see if I can help - I requested access to the files.

Kind regards,
Mateusz

[Mateusz Staniak]

Hi,

one more thing: we can't reproduce the error without your annotation file.

Kind regards,

Mateusz

[Alison Ross]

HI Mateusz, thank you very much for your reply! 

I granted access to the two MaxQuant outputs on Google drive and the annotation file is attached!

Cheers,

Alison 

[Mateusz Staniak]

Hi,

it looks like in the annotation, for Mixture3 there are two condition per channel in some cases. Why?

Kind regards,

Mateusz

[Mateusz Staniak]

Hi,

not sure if it was clear, but that is the source of your problem. Each Channel in a Run should correspond to a unique Condition (can't have two or more conditions per channel).

Kind regards

Mateusz Staniak

[Alison Ross]

Hi Mateusz, 

Thank you very much for your reply! So to be clear, you mean that you cannot group samples of similar types in the annotation file but rather must name each individual sample? Or was this an error specific to Mix 3? I didn't see anything different about Mix 3 compared to the others, but maybe I missed it! 

Is there a way to group samples for downstream analysis? E.g. samples 1-4 corresponding to channels 126, 127N, 127C, 128N are all from condition 1 and I would like to do grouped analyses on these samples to see how conditions 1 and 2 differ from one another.

Cheers,

Alison 

[Mateusz Staniak]

Hi,

unfortunately, this is not what I meant.

What you describe is having multiple biological replicates for each conditions, which is not just possible, it's advised and most welcome.

However, to process the data and prepare for group comparison, we need to know which measurements belong to which groups and replicates. Usually, we expect that in each run, each channel corresponds to a single condition (and single biorep. Conversely, single condition may match to multiple channels due to different bioreps). In some processing steps, we assume that for each  Mixture+TechRepMixture, Channel-Condition annotation is the same across all runs (fractions in your case).

However, in your data, this assumption is not true. For run 20210716_ABR_XX_Mix03_Complex, Channel-Condition pairs are different than in other runs. Because of that, one of processing steps fails. If that Channel-Condition annotation is a mistake, it's enough to fix it. If not, we might need to update MSstats code.

Kind regards,

Mateusz

[Alison Ross]

Hi Mateusz,

Thank you for your clarification, this was a silly error on my part. The MaxQtoMSstatsTMTFormat and proteinSummarization functions are both working now :)

Best wishes,

Alison