Skip to main content
Topic: Adduct-based filtering of featurelist (Read 2325 times) previous topic - next topic

Adduct-based filtering of featurelist

Hello at all,
I´m interested in reducing the number of features of my raw feature list. The steps I´am currently doing are the following:
- MS-Dial data processing: using parameters "peakcount filter" & "N% detected"
- MS-Dial data processing: elimination of blank features (by applying a specific limit value for sample average/blank average ratio)
- manually after data processing: elimination of features by QC-specific filters (using specific limit values for missing rate and CV value)

By performing these steps, I can reduce the number of features from ca. 10.000 (or more) to about let´s say 4000-5000. But there are still redundant features, i.e. adducts!

So my question: Is there a simple way to (semi-)automatically delete all the adducts from the feature list in MS-Dial? How would be a reasonable way to do this?
As I remember right, in XCMS "peak grouping" at least facilitates eliminaton of adducts. In MS-Dial, somehow similar information is given in the column "post curation result" within the aligned feature list, but based on these data you are not able to delete all adducts from the list in a reasonable time.

Unfortunately I´m not experienced in programming. So, I would prefer using a tool that does not require deep knowledge in programming. However, can someone recommend any tool?

One additional question. Just to be sure, the isotopes are excluded automatically from the aligned feature list in MS-Dial?

MS-Dial version: 4.90

Best regards
FalcoB

Re: Adduct-based filtering of featurelist

Reply #1
Hey FalcoB,

The easiest user friendly solution is to use the MSDIAL output directly for MS-FLO here: https://msflo.fiehnlab.ucdavis.edu/#/No programming needed! Also the paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7334838/

Hope this helps and takes care of MOST things you asked!
Thanks,
Biswa