The first law of computing (garbage in, garbage out) applies equally well to screening fragments. Selection of compounds for fragment screening is a theme that I will explore in some depth in future posts. For now, I'll just let folk know that we have already compiled some literature on the topic and that our article on this topic is now available 'online first' at the Journal of Computer Aided Molecular Design. Three of my friends and I put this together as a contribution for the special issue of this journal on FBDD so you can expect to see more articles on the subject appearing soon.
There's going to be some fragment action at the ACS. Mel will be there so look out for her.
Literature cited
Blomberg et al, Design of compound libraries for fragment screening, J. Comput.-Aid. Mol. Des., in press (online first) DOI
1 comment:
This is a nice paper, and provides a particularly clear account of the “Core and Layer” strategy, in which compounds are layered progressively onto a core fragment collection to increase sampling diversity. As I understand it, the core compounds have a higher “perceived attractiveness” than the compounds in the outer layers. What would be interesting to know is how “perceived attractiveness” fares in the context of actual screening. Are the core compounds more frequent hits than the outer layer compounds? And could these data be used to design subsequent libraries?
The paper also provides some practical advice for pooling compounds for X-ray and NMR screening, but again, it would be interesting to see how the libraries fare in the real world. For instance, the X-ray library consists of 10 compounds in 57 groups; how often is the electron density sufficient to unambiguously identify the bound fragment, or, to lift a word from Nino Campobasso, how often does identification come down to a question of blobonomics (interpretation of ill-defined electron density blobs)?
Post a Comment