Page 1 of 1

Data Independent Acquisition (DIA) on first generation Q-Exactive

Posted: Mon Mar 09, 2015 7:01 pm
by karthikskamath
Hi Guys

Did any one in the forum have any experience of testing/running DIA experiments for complex proteomic samples on first generation Q-Exactive? If yes could you please share your experience here? Is the MS/MS scan speed good enough for faster transitions (my peak widths are ~30sec)? I read in one of the older post in this forum that TOF is better system, what is the recent update? Also can some one share the workflow for the DIA data analysis on Thermo system? Is Skyline good enough to analyze the date? I heard there are issue with setting FDR filters??

Posted: Tue Mar 10, 2015 2:13 am
by Zhang
One of the landmark papers is published in Nature Methods, by the group who developed Skyline. Q-Exactive can do DIA, but in a different way than the SWATH on Q-TOF.


karthikskamath wrote:Hi Guys

Did any one in the forum have any experience of testing/running DIA experiments for complex proteomic samples on first generation Q-Exactive? If yes could you please share your experience here? Is the MS/MS scan speed good enough for faster transitions (my peak widths are ~30sec)? I read in one of the older post in this forum that TOF is better system, what is the recent update? Also can some one share the workflow for the DIA data analysis on Thermo system? Is Skyline good enough to analyze the date? I heard there are issue with setting FDR filters??

Posted: Tue Mar 10, 2015 3:33 am
by maccoss
We have a tutorial on how to setup a regular DIA method using Skyline and a Q-Exactive at https://skyline.gs.washington.edu/labkey/wiki/home/software/Skyline/page.view?name=tutorial_dia

Posted: Tue Mar 10, 2015 3:34 am
by maccoss

Posted: Tue Mar 10, 2015 6:51 am
by oBernhardt
We recently published an extensive study using DIA on a first generation Q-Exactive instrument.

Extending the limits of quantitative proteome profiling with data-independent acquisition and application to acetaminophen treated 3D liver microtissues
[[url]http://www.mcponline.org/content/early/2015/02/27/mcp.M114.044305.abstract][/url]

In this publication we used Spectronaut to analyze the data. We could achieve an almost complete library recovery (98 %) containing of ~30.000 precursors and very good CVs (Median CVs < 10%) for both, our technical benchmark experiment and a real biological experiment.

The paper provides detailed information of how to setup your experiment and your method in order to get the best performance out of your instrument.
Spectronaut uses a very solid FDR model and also comes with interference correction and cross run normalization.
But to sum it up, the Q-Exective has excellent DIA capabilities.

If you need more information, please do not hesitate to contact us under info@biognosys.ch

Oli


--
Oliver Bernhardt, MSc.
Biognosys AG

Wagistrasse25
CH-8952 Zurich
Switzerland
bernhardt@biognosys.ch
http://www.biognosys.ch

Posted: Thu Mar 12, 2015 4:40 am
by karthikskamath
Many thanks for the video tutorials and the paper. Both are very descriptive. I will go through them in detail

Cheers
K

Posted: Fri Mar 13, 2015 4:50 am
by tsbatth
Hey cool paper was definitely interesting to read and congratulations. The real time DIA was really cool and it really seems methods which can change parameters on the fly are the future.

I had a question regarding the DIA vs DDA comparison. Specifically you use a ms/ms spectral library generated from the 10 or so DDA runs to match features to your DIA runs. Did you consider also applying the same criteria to the DDA runs ? What I mean by that is using the accurate mass tag (AMT) approach on the MS level since the high resolution and mass accuracy at the MS level could be beneficial in this regard. In Maxquant that option is called "Match between runs" which many people are using since MQ also performs a iRT like function but on a global scale for all the peptides. Could make for a even better comparison , but cool stuff nonetheless.

T

Posted: Fri Mar 13, 2015 1:55 pm
by Rolando
Hi Mr. T

Thank you for your kind words on our publication.
You are correct, in shotgun proteomics, MS1 alignment can be performed (also MaxQuant can do that). I have also tried this on our DDA data set. It increases the data set completeness from about 50% to 65% (compared to 98% in HRM-DIA). But also the coefficients of variation increase, therefore further analysis would be required to assess the performance.
In DIA for peptide precursors not identified, an alignment can also be performed, equally to shotgun proteomics. But it can not only be done on MS1 but also on MS2. Since our data set was over 90% complete, alignment is less necessary.

Best
Roland Bruderer