I planned an experimental setup using 10-plex TMT with 9 conditions and 1 bridge channel to compare 5 replicates.
But since I experienced some run-to-run variation, I was wondering if a better setup would be to run 3 conditions x 3 replicates + 1 bridge channel/MS run. The bridge channel would compare 3 experiments instead of bridging between replicates.
In other words, is it better to have several replicates in one MS run and then use the bridge channel to compare different experiments, or is it better to have 1 replicate per run and use the bridge channel to compare the different replicates?
Dear Mariam,
the following article might help you
doi.org/10.1371/journal.pone.0120115
all the best,
Sami
I can see your point, but the first thing you need to address is why are you having run-to-run variation? The LC-MS setup needs to ensure runs are reproducible. If they are not, then you have a problem with the setup up. This is where you need to fix any issues before you can begin your experiments.
On the run setup, if you put all of your replicates in 1 TMT run they will clearly be more comparable than if they are in different TMT runs, but have you then biased your data? Undoubtedly yes. Is the change you detect due the run-to-run variation or the samples themselves?
Samples should ideally be randomised in a “block” design. I would not put replicates in the same TMT run in this case.
Info on sample randomisation;
[1]B. Burger, M. Vaudel, H. Barsnes, Importance of Block Randomisation When Designing Proteomics Experiments, J. Proteome Res. (2020). https://doi.org/10.1021/acs.jproteome.0c00536.
[2]T. Zhang, M.J. Gaffrey, M.E. Monroe, D.G. Thomas, K.K. Weitz, P.D. Piehowski, V.A. Petyuk, R.J. Moore, B.D. Thrall, W.-J. Qian, Block Design with Common Reference Samples Enables Robust Large-Scale Label-Free Quantitative Proteome Profiling, J. Proteome Res. 19 (2020) 2863–2872. https://doi.org/10.1021/acs.jproteome.0c00310.
[3]M. Krzywinski, N. Altman, Points of significance: Analysis of variance and blocking, Nature Methods. 11 (2014) 699–700. https://doi.org/10.1038/nmeth.3005.
Hi Mariam,
I would recommend that you run biological replicates plus reference (or bridge channel) as one TMT 10 plex experiment. This provides the statistical number that you would need to compare, in one TMT 10 plex, 3 conditions. Between the biological replicates, you should be looking for a high correlation of intensities and if the mixing is good, even between the three conditions..
The variability introduced by labeling and mixing between two TMT 10plexes is significant enough, even if you create a bridge sample that contains all the samples..This is why I prefer to create a TMT 10plex that can stand on its own by providing statistical analysis..
Best,
Hediye.
- Badges
- Science topic
- Similar topics
- Chemistry
- General Biochemistry
- Proteomics