Cross compare DDA/DIA results

[animesh]

aim of the new module

1. what is its aim?
Data consists of E.coli Dilution mixed with standard HeLa  measured in DIA and DDA mode with timsTOF-pro2 instrument. Comparing results from these modes will give us a clarity on what is different and preferably some guidelines for users.

2. what is its added value compared to already available modules?
E.coli Dilution were performed and mixed with standard HeLa in 1:1, 1:2, 1:4, 1:5 and 1:10 HeLa. So we can roughly asses the effect of dilutions in both modes.

full description of the new module

1. What file would be available for download on the module website?
Data is available at PRIDE (https://www.ebi.ac.uk/pride/archive/projects/PXD072581/private Username: reviewer_pxd072581@ebi.ac.uk Password: jARL0st6lW0n)
2. What would users need to do?
Read the blog post on substack to get an idea of what has already been done and expand, maybe MaxQuant/MaxDIA?
3. What would be the type/format of files used as input for the module?
Folders *.d
4. What metrics would be calculated and how?
Overlap between DDA and DIA results, their quantification correlation/scatterplot, CV,...
5. Propose a plot for metrics visualisation
Check https://fuzzylife.substack.com/p/dia-nn-learned-to-search-dda for example the LFQbench style plots

potential external expert(s)

No response

Will you be able to work on the implementation (coding) yourself, with additional help from the ProteoBench maintainers?

• yes
• no

any other information

No response

[mlocardpaulet]

Hi,

thanks for proposing this new module. If I understand well, you would like to compare DIA and DDA data in the same module?
This is outside of the scope of ProteoBench, which is dedicated to the comparison of software tool performance, not the comparison of MS instruments and/or acquisition strategy.

Please let me know if you would like to discuss it further with the team :)

[animesh]

Thanks Marie for checking. I think the data is good enough even for the within comparison of DDA or DIA without moving into inter-comparison space. In fact that is part of the comparison in did in my blog https://fuzzylife.substack.com/p/dia-nn-learned-to-search-dda but only using DIA-NN, next obvious tool to check would be MaxQuan i think?

[mlocardpaulet]

DIA-NN outputs for DDA data are now (or will soon be) compatible with current ProteBench modules, so it will be easy to compare its performances with MaxQuant and other tools. I think that your comparison is really nice but I miss what more it will bring to ProteoBench that is not there already. Since we have DIA and DDA data from different instruments. We don't want to have too many redundant modules, since it risks "diluting" point submission, and it will make it harder to maintain. Any other opinion on this?

[animesh]

Thanks for appreciation @mlocardpaulet 🙏 Yes that is actually the point of this dataset, it is from the same instrument, same samples, run in different modes, DIA and DDA. Thus confounding factors are low 🤞 I think as a researcher biologist are interested in knowing where to use what? Do you have such module comparing difference software, different modes?