FluoroMatch – Covers entire non-targeted PFAS workflow

1. New homologous series voting algorithm alongside formula prediction with a < 5% false positive rate for these predicted formulas!
2. New report covering all parameters
3. Support for .mzML, .mzXML and .ms2, .d, .raw (Thermo and Waters), .wiff, etc. Support for all major vendors (Agilent, Shimadzu, SCIEX, Thermo, Waters, Bruker...)
4. Major debugging and increased stability:
   • loading bar works now
   • BFF (blank filtering) for MZMine 3 works now
   • Formula prediction does not frequently error out
     etc.
5. Ability to close software and terminate all processes (with warning)

We are excited to release the newest version of FluoroMatch ahead of ASMS 2024

FluoroMatch 4.45 updated on 5/9/2024

1. Formula Prediction: Now FluoroMatch has a built-in, unique, and highly sophisticated formula prediction algorithm catered to PFAS. This feature was led by and developed by David Schiessel (see the Figure below for the formula prediction algorithm workflow).
2. New FluoroMatch Visualizations: Updates include the ability to filter by fragment intensity when using fragment screening, display of fragment m/z values in the fragment screening table (scroll right to see), addition of a visualization showing theoretical versus experimental isotopic patterns for predicted formulas, additional flags for users to annotate datasets, and a new layout. These changes were led and developed by Jeremy Koelmel (Research Faculty, Yale University, under Krystal Pollitt’s Research Group) and David Godri (3^rd Floor Solutions).
3. Implementation for Waters .raw data files: See updates to the written manual for instructions on how to use this new implementation. Workflow updates include a new tool for removing lockmass and other scans, optimized peak picking, and other modifications to allow FluoroMatch to work for Waters files. These changes were led and developed by Jeremy Koelmel, Juliane Brown (USGS), and James Gray (USGS).
4. New and modified libraries: including the addition of PAPs and diPAPs (standards acquired on an Agilent Q-TOF), only showing linear not branched hits (since they cannot be discerned) and editing annotation rules to reduce false positives. These changes were led and developed by Jeremy Koelmel, Carrie McDonough (Professor, Carnegie Mellon University), and Leenia Mukhopadhyay (Carnegie Mellon University).
5. Faster script for generation of MS and EIC visualization files: this includes C++ integration with FluoroMatch (Developed by Michael Kummer and Jonathan Sparks).
6. New algorithm for linking MS/MS cans to features: either by ppm mass accuracy or isolation window (divided by 4).
7. Updated the distributed R used by Flow and Modular: updated to R 4.4.0 to overcome security concerns with previous versions of R. Note if you look in FluoroMatch-4.45_GitHub\Flow\ it still shows R-4.2.1, but this is actually R 4.4.0 (we temporarily renamed to folder so the software would still know where to find R. This naming will be fixed in the next release).
8. Inclusion of various MSConvert versions: this is to try and overcome loss of DLL and different vendor support issues.

Formula prediction algorithm workflow

Other Acknowledgements:

Brian Ng (FDA, Maryland), Christine O’Donelle (FDA, Maryland), Juliane Brown (USGS, Colorado), James Gray (USGS, Colorado), Sheng Liu (Krystal Pollitt laboratory, Yale University), Jacqueline Bangma (US EPA), Kathy Hunt (Vogon Labs), Rachel Smolinski (Carrie McDonough laboratory, Carnegie Mellon University), Jake Thompson (John Bowden’s laboratory, University of Florida), Olivier Chevallier (Agilent Technologies) and Stephan Baughman (Agilent Technologies), to name just a few cutting edge researchers for substantial testing of the software on various instrument platforms, providing testing files, and providing in-depth feedback.

Agilent Technologies, for supporting Innovative Omics by providing funding, data, and expertise under the mentorship of Emma Rennie to continue to develop FluoroMatch (including many of the updates above).

FluoroMatch 4.3 updated on 11/09/23

1. Isotopic Support (great thanks to Michael Kummer, and David Schiessel!):
   1. Filter by percent abundance of isotope
   2. Kauffman Plots with reference database
   3. Working on Sophisticated Formula Prediction (coming very soon!)
2. Dramatic Library Expansion (great thanks to Jacqueline Bangma, Mark Strynar, Carrie McDonough, David Dukes, and Sheng Liu):
   1. 8,874 new rule based fragmentation across 29 compound classes (see b and d)
   2. Biotransformation Products from Enzymatically Treated Standards
   3. Biotransformation products using Biotransformer and structure-fragment relationship prediction (200,000+)
   4. Chemical compounds related to industrial effluent characterized at the EPA
3. New filters
   1. Filter by feature number, m/z, retention time range, isotope percentages...
   2. Filter by flagged features
4. IonDecon (great thanks to Michael Kummer, Kathy Hunt (Vogon Labs), and Olivier Chevallier!):
   1. This is a seperate software but now allows FluoroMatch to support All-Ions data: IonDecon (all-ions deconvolution) – Innovative Omics (working to integrate into FluoroMatch to increase throughput, coming soon!)

1. A New Interface with new features (give great thanks to Bernard Brooks!):
   1. Important/relevant weblinks
   2. Loading bars which track the progress of all steps (file conversion, peak picking, annotation, visualization files)
   3. New parameters which can reduce file size and speed up processing (intensity cutoffs for file conversion and annotation, retention time range for file conversion, etc.)
   4. debugging functionality to rerun only the annotation step after peak picking if there was an issue
   5. automatic saving of parameters for next time it is used
2. Statistics integration (great thanks to David Godri and Bernard!):
   1. New "groupings parameter" on the interface for automatic group detection
   2. ANOVA, Volcano Plot, PCA, Violin Plot, Heatmap, etc. calculations and visualizations
3. Targeted peak detection, just upload a peaklist and targeted peak detection will be performed and added to the non-targeted peak table. For example, add your list of standards (m/z and retention times) or your list of knowns you want to make sure to capture can be uploaded every time you run the software. Could also be used for retention time predicted databases for searching.
4. Chromatogram (EIC) and MS1 visualizations (great thanks to Michael Kummer and Nandarani Abril). Visualize chromatograms interactively for every feature in the visualizer along with MS1 spectra with isotopic labeling.
5. Python integration into the FluoroMatch ecosystem (thanks to Eric Siegel, Michael Kummer, and Nandarani Abril)
6. R v4.2.1 integration into the FluoroMatch ecosystem (thanks to Bernard Brooks)
7. New libraries (labeled standards and transformation products):
   1. PFAS transformation products especially glucuronic acids (certain libraries are turned off until publication/discussion, but will be turned on shortly) (great thanks to the Carrie McDonough group, and Sheng Liu of Krystal Pollitt's group)
   2. A subset of labeled standards (thank you to Jackie Bangma and James McCord)