EMP metabolomics

The latest metabolomics data/jobs/results and methods are available from that repository https://github.com/lfnothias/emp/blob/release_2.1/methods/methods_release2.md.

The notebooks below were used to preanalysis and prepare the metabolomics results, including the feature metadata master tables.

The latest feature metadata master tables for Feature-Based Molecular Networking (FBMN) and Classical Molecular Networking (CMN) are available at https://github.com/lfnothias/emp/tree/release_2.1/data/metabolomics/.

Prepare feature metadata

EMP_Metabo_FBMN_CMN_Concat_Annotations.ipynb

This notebooks takes all the annotations available and concatenate them into a master feature metadata table.

Interative notebook here ->

Visualize intensity distribution in FBMN

EMP_Metabo_FBMN_visualize-distrib_v4.ipynb

This notebook visualized the intensity distribution in the metabolomics feature intensity table. That informations helps to define an intensity threshold for noise and for abscence/presence of a compound. Two feature tables are used (before/after gap filling) and for simplicity, we use a subset of samples.

Interative notebook here ->

Interative notebook to compare before/after gap filling here ->

Convert MZmine table to BIOM

EMP_metabo_MZmine_to_BIOM.ipynb

This notebooks takes a MZmine feature table and convert it into the BIOM format.

Interative notebook here ->

Consolidate structures identifiers

EMP_Metabo_Annotations_consolidate_CMN.ipynb EMP_Metabo_Annotations_consolidate_FBMN.ipynb

This notebooks consolidate structure identifiers.

Interative notebook here ->

Interative notebook here ->

Establish microbial metabolites and propagate

EMP-metabo_CMN_microbial_metabolites.ipynb EMP-metabo_FBMN_microbial_metabolites.ipynb

We match putative spectral annotation against NPAtlas and MIBIG and recover metadata (suffix/prefix _NPA_ and _MIBIG_ respectively). We create new columns to indicate if an annotation is microbial (is_microbial). We differenciate the annotation level based on Metabolomics Standard Initiative standards. We also create new columns that store the identifiers from NPAtlas and MiBIG.

- Level 2: GNPS spectral library match in regular and analaogue mode, for `GNPS_LIB` and `GNPS_LIBA`
- Level 3: DEREPLICATOR and DEREPLICATOR+, for  `DEREP` and `DEREP+`
- Level 4: SIRIUS/CSI:FingerID, for `CSI_`

For example is_microbial_level_2 or by combining levels is_microbial_level_2_3_4. 'YES' indicates this metabolites is potentially microbial and belong to the respective annotation level.

is_microbial_tool column summarizes the annotation tool that gave a microbial metabolite hits.

is_microbial_tool_id column summarizes the NPAtlas and MIBIG identifiers for the hits.

• We propagate the microbial molecules using molecular network families (using GNPS_componentindex) and create a new column to indicate that these molecules are part of a putative microbial network. For example with columns: is_microbial_level_2_3_4_network and is_microbial_level_2_3_4_networkid (for id of GNPS_componentindex).

Interative notebook here ->

Interative notebook here ->