Using GNPS FBMN prior to SIRIUS
Closed this issue · 2 comments
Hello,
I am annotating my DDA file using MZmine, FBMN, and SIRIUS.
I found that after extracting features using MZmine, I can directly pipe in to the SIRIUS or I can also first perform molecular networking then put the output for SIRIUS.
I have done both ways and am finding hard to distinguish two results.
Am I missing some information or is molecular networking irrelevant for improving SIRIUS analysis?
Thank you!
Best wishes,
Minsoo
That's an interesting question. I'm using Sirius to help annotate my GNPS networks with molecular formula and predicted compound classes. I hadn't thought if Sirius can use the molecular networking results to help it with determining formulas and classes. They are made by different teams so idk how much they co-ordinate. I know Sirius uses Zodiac to help with formula prediction "by taking similarities of compounds in the dataset into account" but its a different method independent of the GNPS molecular networking I think. I'm just a user not a developer though so I don't know the details of how it works exactly
Hey,
SIRIUS does not use any information from molecular networking to compute its results.
However, if you export MGF,s GNPS and SIRIUS using MZmine, then it is ensured that features ids of the results are compatible and results can mapped to each other.
Slightly related:
In SIRIUS 6, you can add spectral libraries as custom databases to SIRIUS. SIRIUS then also performs library search in the background and annotates the structure database search results with spectral library hits. It also ensures that formula candidates with a good library match will be part the structure database search. This eliminates the step of merging structure db results and spectral library search results from different tools.
See, https://v6.docs.sirius-ms.io/advanced-background-information/#spectral-library-matching-via-custom-databases
Best
Markus