Somedays ago, I found a report about compound prioritization in single-concentration screening.
http://pubs.acs.org/doi/abs/10.1021/acs.jcim.6b00299
The early stage of drug discovery project, we need to screening lots of compounds. Single-concentration assays are often used because of throughput. In this case, inhibition percent or related output are used.
Medicinal chemist need to prioritize compound using these dataset. In my case, I used PEI (inh%/MW) instead of LE or LLE. But I knew that PEI is not correlate LE ( I’ll check paper ASAP )!
The author proposed new index named single-concentration response value scpIC50(Resp) .
scpIC50(Resp) = -log10[ ( 100/Resp-1 ) * 10^-5 ] —- eq(5)
The equation is same of IC50 when Hill slope = 1.0.
I think this is bold thinking, but the index works well.
In the fig2, scLE and LE showed good correlation. And also, scLE can use hit triage. In the fig 3 and 4 indicated that the index would rescue attractive compounds set.
It was interesting for me.
By the way, recently there are a lot of scores for medchem not only biological/ADME data but also in silico data.
Alogp. Clogp, PSA, …i.e., SAscore, CNS MPO, druggability, LE, LLE, LEAT, PEI, BEI, ……
Which score do you use ?
Is life worth living? 