Try rdkit mmpdb to see if suitable

IBM / pychemex

Python library for Cheminformatics ML model explainability

Apache License 2.0

0 stars 1 forks source link

Try rdkit mmpdb to see if suitable #2

Open JJanowiak opened 3 years ago

JJanowiak commented 3 years ago

get rdkit, try mmpdb, see how fragment data is stored / retrieved. see if this is going to be suitable

JJanowiak commented 3 years ago

https://github.com/rdkit/mmpdb

BenjaminTayler-Barrett commented 3 years ago

A/C:

mmpdb should be used to produce a SQLite database
A method to produce this SQLite database should be documented in a common place

BenjaminTayler-Barrett commented 3 years ago

At present, can get fragments I think...

They do not seem to present anywhere in the resulting database. If the fragments that are generated are sensible, it should be fairly trivial to produce our own database from the fragment files that are produced by mmpdb.

I think this warrants further discussion. @JJanowiak will need to verify that the fragments are sensible though. We should discuss at next meeting. If useful, will document the process used to get to this point.

Alex-AMC commented 3 years ago

An initial set of fragments have been calculated with RDkit. We can come back to this or look to expand the current fragments (85) using this tool. @BenjaminTayler-Barrett can you get those fragments ripped out of the data set and into a sensible format? I won't close this for now as the RDkit fragments are a lot simpler.