CompoundsIdentifiers

This repository provides tables of metabolites identifiers.

The tables folder is the main folder and contains the following tables:

• HMDB-to-KEGG_<date:MM-DD-YY>.tsv: table with HMDB identifiers matching to KEGG identifiers.
• LipidMaps-to-KEGG_<date:MM-DD-YY>.tsv: table with LipidMaps identifiers matching to KEGG identifiers.

---

A nice compound

---

Further details on how the tables were produced

Re-producing the tables is completely optional.

Expected structure of this folder for re-producing the tables:

├── extra
│   └── <empty>     <- here the hmdb files (see "Instructions for HMDB")
├── scripts
│   ├── hmdb.py
│   ├── hmdb_ids-tokegg.py
│   └── lipidmaps_ids-tokegg.py
└── tables
    └── <empty>

The scripts in scripts folder produce the content of the tables folder:

• hmdb_ids-tokegg.py : HMDB-to-KEGG_<date>.tsv
• lipidmaps_ids-tokegg.py : LipidMaps-to-KEGG_<date>.tsv

The extra folder contains HMDB related data required to generate HMDB-to-KEGG_<date>.tsv; for more information see "Description of the scripts" section.

Description of the scripts

• lipidmaps_ids-tokegg.py: searches with the API of LipidMaps database and directly generates the final table. It takes 1 minute.

• hmdb_ids-tokegg.py and hmdb.py: Detailed information, and/or if needing to re-run it, see the Instructions for HMDB. Importantly hmdb.py is an identical copy of the fantastic script created by yufree (https://gist.github.com/yufree/f552d865096010445fc7b969e7e9d439) which is an xml parser suited for the hmdb dowloaded database; takes ~ 10 min to complete.

Instructions for HMDB

If willing to re-run the scripts to produce the final HMDB table, before anything, the following is required:

1. go to https://hmdb.ca/downloads
2. download "All Metabolites" dataset, it's a zip file
3. locate the zip file into the extra/ folder
4. unzip the file, this will produce the hmdb_metabolites.xml file, of ~6.5GB (too heavy, so not included in this repository)
5. The two scripts must be executed in the following order:
   • a) Run hmdb.py and wait until completed (~10min); it outputs extra/hmdb.csv
   • b) Run hmdb_ids-tokegg.py, it uses the output of a); it outputs tables/hmdb_ids-tokegg.py

Note : the generated table provided in this repository used the HMDB Version 5.0 Released on 2021-11-17 downloaded on 07-May-2024 .

Help

Contact johaGL for any question