AutoQC suggestions

support
AutoQC suggestions Alejandro Cohen  2018-05-31 10:41:12
 

Hi Panorama people,

I'm using AutoQC and Loader, and I am marveled by it.... a long overdue solution to QC in MS laboratories. I'm using it with an aliquoted BSA digest using MS1 filtering on Skyline (VeloPro acquisition) and this has given me a great tool to assess the overall LC-MS performance prior to running real samples, KUDOS to ALL your team!!!

Now my suggestions:

-Is it possible to add to the QC plots types an additional parameter such as Xcorr or some other output parameter to assess the MSMS quality for DDA acquisitions?? This would be great!!! You already now the sequence of the target peptides, and it is likely you can pull out the MSMS for that precursor in each file. If you could monitor the additional parameter to follow over time, it would be a great source to assess the 'health' of the MSMS.

-Have you considered selling/distributing a QC sample (protein digest?) to be used as an external QC like done typically in clinical labs? It would be great to not only have internal QCs, but also external QCs and evaluate how your instrument is performing compared to identical (or not) instruments in other labs (anonymous). I imagine the huge variability as everyone is using different LC conditions, but if this could be minimally standardized, it could be a source of precious information to evaluate inter-laboratory variability

Cheers

Alex

 
 
Mike MacCoss responded:  2018-05-31 11:05:06

Hi Alex,
Great questions! Let me tell you a bit how we use this in our lab.

The first thing we do is perform targeted MS/MS (aka PRM) on a set of target peptides in addition to MS1. This does a couple of things. 1) It confirms the quality of the MS/MS data in addition to the MS1 data and 2) you feel more comfortable that you are measuring the target peptide analytes that you think you are measuring. We have had cases where we are getting good MS1 data but ran out of something like CID gas so the MS2 data failed. The more you can make this a targeted workflow the better you will be. Doing PRM or SRM on your target peptides works on any instrument platform from your Velos, to a QQQ, or a Q-TOF/Q-Orbitrap etc...

The answer to the second question is that we tend to use Thermo's PRTC peptide mixture as a system suitability standard when we are checking the instrument performance. To confirm digestion QC and perform within sample QC we tend to use CIL's 15N-ApoA1 for human samples http://shop.isotope.com/productdetails.aspx?itemno=NLM-9539-PK and normal unlabeled ApoA1 for non-human samples.

Let us know if you have any other questions,
Mike

 
Tobias Kockmann responded:  2018-07-06 03:58:45

Hi Alex,

This might be of interest. Core4life (https://coreforlife.eu/) is developing such a std. sample in collaboration with Promega and a corresponding analysis pipeline named QCloud:

Chiva, C., Olivella, R., Borràs, E., Espadas, G., Pastor, O., Solé, A., & Sabido, E. (2018). QCloud: A cloud-based quality control system for mass spectrometry-based proteomics laboratories. PloS One, 13(1), e0189209. http://doi.org/10.1371/journal.pone.0189209

In the future, QCloud will be used to compare instrument performance across core labs in Europe. Unfortunately, there is no public access yet, but I think the std. sample will become available soon through Promega. Eduardo Sabido and Karl Mechtler should be able to provide further details.

Greetings,
Tobi

 
Alejandro Cohen responded:  2018-07-11 07:35:34

Thanks Mike and Tobi for your feedback! Very helpful. I will follow-up with the QCloud people. Eventually, it would be a great component to AutoQC.
Cheers
Alex

 
Tobias Kockmann responded:  2018-07-11 07:43:07

It would def. be cool to also allow Panorama users to compare their instrument performance with others (across projects). Not sure if that is easily doable in the lab key framework.

 
Brendan MacLean responded:  2018-07-11 09:00:01

I am sure it would mostly be an issue of adding a way for users to indicate they are willing to share publicly. Interesting idea. We will certainly discuss it.

 
Alejandro Cohen responded:  2018-07-11 11:25:55

Going through the QCloud paper, I notice some overlap with Panorama's AutoQC. My experience is that the 'survival' and success of this type of platforms relies on the user-base, and I think AutoQC (with Skyline/Panorama's user-base and support network) has an edge here. However, it would be nice to see if AutoQC could 'borrow' some of the QC tools shown in QCloud. The adoption of a 'standarized' sample needs to be probably discussed among lab and core facility users in one of ASMS's Skyline meetings, or the workshops. My two cents.

 
Tobias Kockmann responded:  2018-07-11 12:14:56

Hi Alejandro,

I abs. agree! Both pipelines a more or less complementary. Skyline + Panorama comes from a targeted MS background, while QCloud was clearly developed with a shotgun philosophy in mind. Both strategies have their pros and cons and we tend to use both to monitor instrument performance on different time intervals. From a tech. perspective, Panorama is def. in the leading position.

Greetings,
Tobi