All Notebooks | Help | Support | About
Archives
Authors
Sections
Tools
Show/Hide Keys
28th May 2013 @ 13:39

Paul Ylioja applied the SMARTS filter

http://pasilla.health.unm.edu/tomcat/biocomp/smartsfilter

(using "expert mode") to the ChEMBL-NTD database. Output is below as a zip file containing 27 files, each addressing 1000 compounds. Files show the SMILES for each compound and the pass/fail outcome according to each criterion.

Analysis was done in June 2012, in consultation with Mat Todd, Mike Robins and Murray Robertson.

Relevant, related literature: http://dx.doi.org/10.1016/j.drudis.2010.08.010

Attached Files
16th April 2012 @ 11:06
A cityscape map was created showing compounds that are similar to two compounds, TCMDC 135294 and TCMDC 134395 in same way as before
purchasable chemical space map centered on OSDD synthesised compounds


In the map, compounds are represented as nodes and connected to another node by an edge if the similarity between them is >0.3 using ECFP_4 fingerprints.
Nodes in the map are coloured by ALogP values and labelled with E-molecules ID
Attached Files
14th April 2012 @ 15:18
Cytoscape maps were generated based on transformations for compounds synthesized using both the Classic transformation and database transformation components in Pipeline Pilot (A commercial product).

Two reports were generated focusing on just ZYH-3-1 child compounds that would have a ALogP<5 resulting from both types of transformations. There is also a pdf showing a map centered on ZYH-3-1

In the maps,
Node ID shows the type of transformation
Node size reflects the ALogP of the compound, larger nodes have LOWER values. The smallest nodes have ALOGP >5
Node colour reflects the change in AlogP of the child compound vs the original. Red nodes indicate in a drop of >1, Green nodes a change of >0.5 while Blue nodes indicate an INCREASE ALogP
Attached Files
11th April 2012 @ 15:47
A purchasable chemical similarity map centered on compounds synthesised, list obtained from [url= http://bit.ly/OSDDcompounds]http://bit.ly/OSDDcompounds[/url]

Map is generated as before, except two thresholds were used. 0.7 for very similar compounds, and 0.5 for similar compounds. The lower the threshold the more connections and nodes in the network.

A small network is also shown which contains all the neighbours of ZYH-3-1 from the similar map. This can be generated by selecting the node of interest (search box is useful), choosing "Select -> Nodes -> First neighbours of Selected Nodes", then "File -> New Network -> From Selected Nodes All Edges"
Linked Posts
Attached Files
11th April 2012 @ 09:03
A structural similarity map of compounds from the publication grouping GSK's compounds into high priority compound lead series (http://pubs.acs.org/doi/abs/10.1021/ml200135p) including very close purchasable structural neighbours.

The Fast track publication contains ~550 compounds grouped into ~40 compound series' for follow up by the community. As with previous purchasable similarity map, original compounds are shown as red nodes and purchasable compounds from E-Molecules as blue nodes. Nodes are connected by an edge if compounds have a structural similarity of > 0.7 Tanimoto co-efficient with ECFP_4 fingerprints (i.e very similar compounds)

The map is best viewed with Cytoscape (http://www.cytoscape.org/) and the ChemViz plugin (http://www.cgl.ucsf.edu/cytoscape/chemViz/). Nodes can be right clicked to view compound structure. IDs and smiles strings can also be copied out of Cytoscape in to a text editor.

The E-molecule compounds can be retrieved from the E-molecules list view (http://www.emolecules.com/cgi-bin/rene/list_search_setup.cgi), selecting E-molecules ID from the drop down and inputing the list of ids to be retrieved
Attached Files