gmm
commented
4 years ago Hi Dana,
There is also SuCOS, developed by one of my former DPhil students, Susan Leung, which is an RDKit-based shape+pharmacophoric feature-based overlap score, open source (MIT license) and available from:
https://github.com/susanhleung/SuCOS
Garrett
On 28 May 2020, at 3:27 pm, Dana Klug notifications@github.com wrote:
As we move forward with the MurE project, we are seeing the need to look at the original hit fragments in a new way, since screening closely matched analogues did not turn up any new binders ( #16 ).
One potential approach is to treat the fragments not as starting points for direct elaboration (as we might in a hit-to-lead project) but as tool molecules which have revealed available binding interactions in an accessible part of the protein. The question is whether there is a way to use this information to generate new scaffolds and synthetic targets, perhaps by looking at 3D shape and electrostatic similarity to the original hits. The OpenEye program ROCS does something like this and we could potentially apply for their free public domain research license. Ideally we'd like to scan for compounds making use of available interactions that are either 1) commercially available, 2) similar to known compounds in the lit, or 3) requiring exploration of novel chemical space.
@drc007 Any thoughts on open source alternatives or other approaches, or people we could loop in here (either GH or Twitter)? John Chodera's team?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
danaklug
As we move forward with the MurE project, we are seeing the need to look at the original hit fragments in a new way, since screening closely matched analogues did not turn up any new binders ( #16 ).
One potential approach is to treat the fragments not as starting points for direct elaboration (as we might in a hit-to-lead project) but as tool molecules which have revealed available binding interactions in an accessible part of the protein. The question is whether there is a way to use this information to generate new scaffolds and synthetic targets, perhaps by looking at 3D shape and electrostatic similarity to the original hits. The OpenEye program ROCS does something like this and we could potentially apply for their free public domain research license. Ideally we'd like to scan for compounds making use of available interactions that are either 1) commercially available, 2) similar to known compounds in the lit, or 3) requiring exploration of novel chemical space.
@drc007 Any thoughts on open source alternatives or other approaches, or people we could loop in here (either GH or Twitter)? John Chodera's team?