Opened 5 years ago

Last modified 5 years ago

#3473 accepted enhancement

Plugin to Ligand Expo

Reported by: Tristan Croll Owned by: Eric Pettersen
Priority: minor Milestone:
Component: Structure Analysis Version:
Keywords: Cc:
Blocked By: Blocking:
Notify when closed: Platform: all
Project: ChimeraX

Description

Would be really nice to have an interface to the Ligand-Expo (and/or PDBeChem) engine for searching the Chemical Components Dictionary. In particular, they have a substructure search option that could be extremely useful - currently their web interface requires the user to sketch the substructure they want in an applet, but if the query could be generated straight from an atom selection it would be *so* much easier/faster.

Change History (7)

comment:1 by Eric Pettersen, 5 years ago

Status: assignedaccepted

So envisioning this as an API (rather than some sort of direct user tool)?

in reply to:  2 ; comment:2 by Tristan Croll, 5 years ago

I think it would be nice for there to be a user tool as well. Something 
that echoes the standard search fields in their web interface, plus adds 
the search-by-substructure ability based on an atom selection (would 
have to think a bit about how to help the user define the selection - a 
fragment of a single residue - to search by).

On 2020-07-07 19:57, ChimeraX wrote:

comment:3 by Eric Pettersen, 5 years ago

I'm still pretty hazy about what the user would be doing with the search results.

in reply to:  4 ; comment:4 by Tristan Croll, 5 years ago

My immediate application would (surprise, surprise) be about model 
building. Say you have density that is pretty clearly a lipid (the "head 
and shoulders" appearance is pretty distinctive) but the resolution 
isn't good enough to tell you on its own what the head group could be 
(choline, ethanolamine, serine, etc., etc.). The idea is that the user 
would be able to select the "certain" component of the residue (the 
atoms they're sure about) and say "find me all ligands in the CCD 
containing this substructure" so they can choose the one that's most 
sensible for the chemical environment.

I can also imagine it being interesting for people into drug design: 
select the core "scaffold" atoms of a ligand, then fetch and align all 
other ligands based on that scaffold to generate potential ideas for 
elaboration.

On 2020-07-08 18:45, ChimeraX wrote:

comment:5 by Eric Pettersen, 5 years ago

Where is the documentation for web-service access to these things?

in reply to:  6 ; comment:6 by Tristan Croll, 5 years ago

Instructions on how to search by SMILES are at the bottom of here: 
https://www.rcsb.org/pdb/software/rest.do.

On 2020-07-09 18:27, ChimeraX wrote:

comment:7 by Eric Pettersen, 5 years ago

Priority: moderateminor

So this would require converting an arbitrary substructure into its equivalent SMILES string? That isn't trivial. I think OpenBabel could do it, but I don't know how robust/reliable it is for that. And it would be mean installing OpenBabel. Unless you know another alternative...

Note: See TracTickets for help on using tickets.