search

skschum / MFAssignR

The MFAssignR package was designed for multi-element molecular formula (MF) assignment of ultrahigh resolution mass spectrometry measurements. A number of tools for internal mass recalibration, MF assignment, signal-to-noise evaluation, and unambiguous formula selections are provided.
GNU General Public License v2.0
22 stars 24 forks source link

Unable to detect iodine-containing compounds while bromine is detected in high-resolution mass spectrometry data using MFAssignR #32

Open T0mohiro0 opened 1 month ago

T0mohiro0 commented 1 month ago

I am encountering an issue with the detection of iodine-containing compounds using MFAssignR. Despite the presence of iodine in my samples, the library fails to identify any iodine-containing molecular formulas. Interestingly, bromine-containing compounds are successfully detected. Here are the details of my problem:

Description of the issue:

Steps to reproduce:

  1. Load high-resolution mass spectrometry data (m/z range: 143-528).
  2. Run MFAssign or MFAssignCHO with the following parameters:
    • ionMode = "neg"
    • lowMW = 140
    • highMW = 530
    • Ix = 5
    • IAtomEx = TRUE
  3. Check the output for iodine-containing formulas.

Expected result: Detection of iodine-containing compounds, including those in the form of IOx, similar to the successful detection of bromine-containing compounds.

Actual result: No iodine-containing formulas are detected in the output, while bromine-containing formulas are correctly identified.

Environment:

Additional information:

Could you please advise on how to modify the library or its usage to properly detect iodine-containing compounds, given that bromine detection is working correctly? I would be happy to provide any additional information or test data if needed.

Thank you for your time and assistance.

Best regards, Tomo

skschum commented 1 month ago

Hello Tomo,

I am glad you are interested in MFAssignR and I hope it can be useful for you.

For the iodine molecules, are they in any kind of series (CH2, H2O, etc.) with other lower mass (below DeNovo) iodine molecules, or are they isolated compounds without any real connection to other iodine molecules? MFAssignR in general relies on homologous series to assign higher mass molecules, exploiting the general molecular relationships in complex organic matter, which works great for a lot of data, but if a molecule in not related to anything else (for example a spiked standard) it can be harder to assign.

However there are some parameters that can be changed to get isolated assignments. The two terms are MSMS and DeNovo. MSMS allows you to try and directly assign every ion mass below the DeNovo threshold, which can help getting assignments. The DeNovo threshold determines the mass threshold where masses are no longer assigned directly and only by formula extension. The default threshold in MFAssignR is 300, but it can be increased, particularly if you are using higher resolution instruments like FT-ICR.

My recommendation is to set MSMS = "on", and DeNovo = 450 or so and see if that helps at all with assigning the iodine molecules.You can also try setting Ambig = "on", though this does lead to many assignments for each mass that you will have to filter through to determine which assignment is best.

If none of this helps, please let me know and I can check on iodine again to make sure it is working properly, or think about any other ideas I may have to make it work. If you have any other questions, please let me know.

Good luck with the analysis and I hope it works for you.

Thanks, Simeon

On Sat, Aug 3, 2024 at 5:13 AM T0mohiro0 @.***> wrote:

I am encountering an issue with the detection of iodine-containing compounds using MFAssignR. Despite the presence of iodine in my samples, the library fails to identify any iodine-containing molecular formulas. Interestingly, bromine-containing compounds are successfully detected. Here are the details of my problem:

Description of the issue:

  • The sample contains known iodine compounds (e.g., C13H8IO7 with m/z ≈ 402).
  • MFAssignR does not detect any iodine-containing formulas in the output.
  • Bromine-containing compounds are successfully detected and identified.
  • The issue persists even after adjusting parameters such as Ix and IAtomEx.

Steps to reproduce:

  1. Load high-resolution mass spectrometry data (m/z range: 143-528).
  2. Run MFAssign or MFAssignCHO with the following parameters:
    • ionMode = "neg"
    • lowMW = 140
    • highMW = 530
    • Ix = 5
    • IAtomEx = TRUE
  3. Check the output for iodine-containing formulas.

Expected result: Detection of iodine-containing compounds, including those in the form of IOx, similar to the successful detection of bromine-containing compounds.

Actual result: No iodine-containing formulas are detected in the output, while bromine-containing formulas are correctly identified.

Environment:

  • RStudio version: 2024.04.2 Build 764
  • Operating System: Windows 11

Additional information:

  • The data is from FT-ICR MS measurements.
  • Other non-iodine compounds, including bromine-containing ones, are correctly identified.
  • I have tried adjusting various parameters, including mass error tolerance and SNR, without success.

Could you please advise on how to modify the library or its usage to properly detect iodine-containing compounds, given that bromine detection is working correctly? I would be happy to provide any additional information or test data if needed.

Thank you for your time and assistance.

Best regards, Tomo

— Reply to this email directly, view it on GitHub https://github.com/skschum/MFAssignR/issues/32, or unsubscribe https://github.com/notifications/unsubscribe-auth/AF72HFCTBOLRZ3DIE4ZBBA3ZPSUONAVCNFSM6AAAAABL5ZHNX2VHI2DSMVQWIX3LMV43ASLTON2WKOZSGQ2DMMRXHEYTKNI . You are receiving this because you are subscribed to this thread.Message ID: @.***>

T0mohiro0 commented 1 month ago

Hello Simeon,

Thank you for your prompt and detailed response.

For this experiment, I am using NH2Cl as the oxidizing agent. I’ve been trying various approaches to detect, but I haven’t had much success so far. I’ll give it another try based on your hints

Thank you again for your support!

Best, Tomo