Retrorules sqlite is not available

[dhananyashunmugam]

I was trying to run MEANtools using the RetroRules database, but it seems the SQLite file is not available.

I have checked the RetroRules website, Zenodo and GitHub releases, but I could only find the TSV files rr02, rr01. The script mentions a SQLite database (mvc.db) for RetroRules, which I could not locate.

Could you please advise if the SQLite file is available somewhere or if there is an alternative way to use MEANtools without it?

[plantomics]

Hi,

RetroRules has recently updated the database and I believe they have stopped supporting the availability of mvc.db. With the old version, we had converted the mvc.db into MEANtools compatible files. These are:

1. Combined_small_middle_big_datasets (available on Zenodo)
2. MassTransitions.csv (used in the pathMassTransitions.py step)
3. base_rules.csv (Used in the heraldPathways.py step)
4. ValidateRulesWithOrigins.csv (Used in the heraldPathways.py step)

Initially these files were part of the GitHub repository but eventually removed because of the space constraints.

I am providing you the link to download the bottom three files:
https://drive.google.com/drive/folders/1I_songM0Lc_Eh66Wq3FydTL25SIHTUSH?usp=sharing

I hope this helps!

Kumar

[dhananyashunmugam]

Thank you very much for sharing the files and the Google Drive link — this is very helpful.

I had one additional question regarding the correlation step (corrMultiomics.py) in MEANtools. The default decay rates used are 5, 10, 25, and 50.

Could you please let me know based on what these specific four values were selected?
Is it acceptable to test other sequential values (for example 15, 20, 30, etc.)?
How would you recommend selecting the best decay rate to proceed with for the next step (merge_clusters.py)? Is there a specific metric that should guide the choice?

Also i would like to get annotation for SMARTS id Eg: RR00162614 , could you please guide me how to get that?

Thanks

[plantomics]

In principle, the decay rate can take a wide range of values. The figure below can help you choose a sensible setting. Lower decay rates tend to produce more compact modules with tighter associations but fewer features, whereas higher decay rates yield larger, looser modules with broader associations. For additional guidance, see the study linked here: https://pubmed.ncbi.nlm.nih.gov/28408660/