Deciding Which Commercially Available Analogues to Buy Within the Thienopyrimidine Series

24 Aug
Published by JimCronshaw


Request for Help

Analogues of the amino thienopyrimidine lead compound are needed to assess structure-activity relationship's (SAR's) of the compound. To acheive this, both novel synthetic compounds and commercially available compounds must be obtained and submitted for biological screening. The novel synthetic compounds are being synthesised in the Todd lab at the University of Sydney. The commercially available compounds were identified by performing a similarity search on eMolecules using a similarity (Tanimoto) co-efficient of 0.55. The search returned a number of compounds, and a large fraction of these compounds were available from enamine. These molecules are shown below with LogP, TPSA and molecular weight values included (calculated in ChemBioDraw). Also included in the image below are molecular identifiers (CTP = Commercial ThienoPyrimidines) to facilitate identification.

At this stage a smaller number of these compounds need to be singled out. The molecules singled out will ideally possess properties that are most interesting for a biological screen for anti-malarial activity. What this means is largely a function of what input this selection process receives, since this process is very new to me.
Jim Cronshaw


MatTodd's picture

Interesting problem. If you gradually ramp up the Tanimoto, what happens? The obvious thing to do I guess is to just keep the compounds that contain the thienopyrimidines, but does that mean we're being too conservative?

JimCronshaw's picture

The Tanimoto scores are recorded on eMolecules in my browser, but when I check the table of compounds that I saved to eMolecules (here) the Tanimoto scores have vanished from the table.
Because of that I can't show you a link, but it is easy enough to whittle the list down. For instance, the list of 18 original hits (13 of which are from Enamine) can be shrunk down to a list of 13 compounds if the Tanimoto score is raised to 0.56. With a Tanimoto score of 0.58 the list is shrunk to 10, and so on.
Five compounds are available from Maybridge. Should we look a these as well? Two of the compounds are in the top three compounds when arranged by Tanimoto score.
Chemically speaking the aminothienopyrimidine core is no longer present in some compounds once the Tanimoto score reaches about 0.57, meaning that there' a good 10+ compounds in the list with that moiety present in their structures. Those compounds also have relativel high Tanimoto scores, compared to the others.

All look valid suggestions, if resources are limited I would priotitise as follows 1) thienopyrimidines, 2) quinazolines, 3) phenyls

JimCronshaw's picture

Sounds good. Would it be wise to purchase at least one from each of those broad categories so that we have a control (or two) when it comes time for testing?

MatTodd's picture

So we need to order these molecules. It seems to me that we could remove one of 3 or 9, given the level of similarity. We could also do a >4 logP filter and take out compounds 2, 5 and 13, but keep 10 and 12 for their obvious similarity. So that would leave 1, 4, 6, 7, 8, 9, 10, 11, 12 for the order.


Any objections?

JimCronshaw's picture

Sounds good. Compound 5 also bears some resemblance to 9 and 3, so eliminating that choice seems reasonable.

MatTodd's picture

So we received the nine commercial thienopyrimidines from Molport - they arrived in excellent time at Sydney airport (on Mon Oct 22nd) but were then stuck in the airport until yesterday (Oct 26th) for some internal airport reason. I drove the compounds to the airport this morning to send to Vicky Avery's lab. The structures are below, now with OSM numbers since they have become part of the project. The picture and biological data will go here.

Link to the project wiki page for this series.