Small molecule retention time assignment problem

support
Small molecule retention time assignment problem george dubay  2018-05-08 12:22
 
Good Afternoon,
I am using Skyline to evaluate my data. There seems to be a problem with the way it is handling the assignments of the peaks that are isomeric, even though the retention times submitted are in close agreement with the observed times. We would love to see the system assign these peaks automatically without any curation.
Notes of significance:
1.    The chromatography peaks are reproducible within +/1 0.02 minutes for all compounds.
2.    The assigned RT’s are always the best match for the observed RT’s.
3.    The intensity of the peaks makes little difference.
4.    The signal to noise is generally very good.
5.    Another anomaly that I am encountering is:
i.    group of samples consisting of my calibration curve and quality controls is imported to Skyline and curated to match retention times and created the curve things look good.
ii.    A second set of data for samples from plasma and fecal material is added.
iii.    When the added data goes into the system the RT assignments for the initial work are lost. This is easily observed for 2-methyl butyric acid, but others suffered the same loss of the RT assignment values.
iv.    As a result, the data for several of the compound requires reassignment of the retention times.
Please, look the data over and see what you can tell me about the assignments.
Thanks,
George
Data specifics:
1.    All data is coming from a Waters Xevo TS-Q.
2.    The SIS used is 13C6.
3.    The fatty acids (derivatives) fragment in the negative ion mode so the [M-H]- ion yields product ions from the fragmentation.
4.    The sample id # are related as given in the spreadsheet attached(page 3).
5.    The transition list used is in the first window of the Excel file attached.
6. I am working with Will Thompson here at the Duke GCB.
7. Raw data is in the folder attached as well.
8. Samples are run with the standard Waters MassLynx software.
 
 
Brian Pratt responded:  2018-05-08 12:43
Hi George,

Thanks for providing data - I will have a look at this.

Brian Pratt
 
Nick Shulman responded:  2018-05-08 15:38
Skyline is getting confused by the fact that three of your molecules all have the same chromatogram "242 -> 143".

Skyline recognizes this, and does the peak picking separately for each of those three molecules. For each molecule, Skyline comes up with a different ranking in terms of which is the best peak. However, when Skyline actually saves the list of peaks, there is no information which tells Skyline which is the set of peaks goes with which molecule.

This looks like a bug in Skyline. I am going to ask around and see what we can do about fixing it.