Metabolomics Society Forum

Software => MS-DIAL => Topic started by: BBonnefille on April 07, 2021, 02:43:22 AM

Title: Not all my internal standards are detected
Post by: BBonnefille on April 07, 2021, 02:43:22 AM
Hello,

I have two internal standards missing in my data treatment results, despite several tries of data processing using recommended parameters and several tests using optimised parameters more in accordance with my LC MS method (we have an Orbitrap). They are not only not identified: the masses are not found.

The other IS (8/10) are found.
I checked the raw data and the missing IS are detected with high intensities (>10E7) and have good peak shapes.

I thought it was a peak detection problem, so I tried:
- several minimum peak height values (from 30,000 to 90,000)
- different mass slice width (0.05 to 0.15)
- the minimum peak width (5 to 15 scans)
But it did not solved the problem.

I also did some test refining the data collection parameters, and alignment parameters, in combination with the peak detection parameters tested earlier, but nothing solved my problem.

Do you have any idea of what could be the problem?
Thank you for your help.
Title: Re: Not all my internal standards are detected
Post by: Hiroshi Tsugawa on April 13, 2021, 07:46:56 AM
Please send me one of your data files with the information of retention time and m/z list of your standards.
Thanks,
Hiroshi
Title: Re: Not all my internal standards are detected
Post by: BBonnefille on April 13, 2021, 09:59:41 AM
Hi Hiroshi,

Here is a link to download two of my .raw data and an excel file with the information you requested. I have the same problem in ESI negative, but with more IS not picked up by MS DIAL (7/24 not found), so I attached a .raw file in both modes:
grosfi.ch/adm4hAq6bpD

Thank you for your help,
Bénilde
Title: Re: Not all my internal standards are detected
Post by: Hiroshi Tsugawa on April 13, 2021, 07:51:17 PM
Hi,

I just checked your positive ion mode data. It works fine.
Of course, several IS compounds were not detected owing to the structure property.
However, at least, I could detect the peaks that you wrote as "Not detected by MS DIAL in ESI pos" in the excel.

I also pasted the parameters that I used. Thanks,

Hiroshi
MS-DIAL ver. 4.60-dev

#Project
MS1 Data type   Profile
MS2 Data type   Profile
Ion mode   Positive
Target   Metablomics
Mode   ddMSMS

#Data collection parameters
Retention time begin   0
Retention time end   100
Mass range begin   0
Mass range end   2000
MS2 mass range begin   0
MS2 mass range end   2000

#Centroid parameters
MS1 tolerance   0.01
MS2 tolerance   0.025

#Isotope recognition
Maximum charged number   2

#Data processing
Number of threads   1

#Peak detection parameters
Smoothing method   LinearWeightedMovingAverage
Smoothing level   3
Minimum peak width   5
Minimum peak height   10000

#Peak spotting parameters
Mass slice width   0.1
Exclusion mass list (mass & tolerance)

#Deconvolution parameters
Sigma window value   0.5
MS2Dec amplitude cut off   0
Exclude after precursor   True
Keep isotope until   0.5
Keep original precursor isotopes   False

#MSP file and MS/MS identification setting
MSP file   
Retention time tolerance   100
Accurate mass tolerance (MS1)   0.01
Accurate mass tolerance (MS2)   0.05
Identification score cut off   80
Using retention time for scoring   False
Using retention time for filtering   False

#Text file and post identification (retention time and accurate mass based) setting
Text file   E:\0_SourceCode\BugReports\20210414_GrosFich\pos_mzrt_lib.txt
Retention time tolerance   0.5
Accurate mass tolerance   0.01
Identification score cut off   85

#Advanced setting for identification
Relative abundance cut off   0
Top candidate report   False

#Adduct ion setting
[M+H]+

#Alignment parameters setting
Reference file   E:\0_SourceCode\BugReports\20210414_GrosFich\210322_SWS_QC4_ESIpos_14.raw
Retention time tolerance   0.05
MS1 tolerance   0.015
Retention time factor   0.5
MS1 factor   0.5
Peak count filter   0
N% detected in at least one group   0
Remove feature based on peak height fold-change   False
Sample max / blank average   5
Sample average / blank average   5
Keep identified and annotated metabolites   True
Keep removable features and assign the tag for checking   True
Gap filling by compulsion   True

#Tracking of isotope labels
Tracking of isotopic labels   FALSE

#Ion mobility
Ion mobility data   FALSE

Title: Re: Not all my internal standards are detected
Post by: BBonnefille on April 13, 2021, 11:00:20 PM
Hi Hiroshi,

Thanks a lot, I will make a new try using the parameters you used.

Bénilde
Title: Re: Not all my internal standards are detected
Post by: BBonnefille on April 14, 2021, 10:43:01 AM
Hi Hiroshi,

I did all my samples data processing following the exact parameters you gave me, but I still have problems with my IS:
- One of my IS is still not found by MS-DIAL (one of those that were missing previously)
- The other detected by MS-DIAL, which is better. But the identification using a .txt file as in-house library did not work, and the feature in the ion table present a high mass deviation compared to my IS mass (approx 13 ppm instead of < 5 ppm when I am checking with XCalibur).

One of the group I am processing a few samples were not spiked with the IS(3 vs 28 samples spiked). Do you think it can be related (even if it should not)?

Thank you for your help,
Bénilde