Open mattodd opened 3 years ago
@mattodd @wwjvdsande @fantasy121 @MA-Jjingyi @MFernflower here is the graph on Metabolic clearance versus logD. The analysis is based on 434 compounds from the Epichem library. There are 927 entries in the library, but most of them are missing either calculated logD values or clearance data. Therefore, the graph was generated based on compounds having full data set.
Pharmacokinetic data suggests that in contrast to our defined lipophilicity threshold value of ≤2.5, molecules with logD ranging from approximately 3.3-5.2 have the highest metabolic clearance, though some outliers are present as well (eg: logD 2.6, 332 μL/min/mg; logD 6.1, 217 μL/min/mg).
For what it's worth the azoles (of which our fenarimol derivatives can be thought of as a sister class to) have variable half lives but as a general rule it's said to be around 20 hours on the high side and 16 on the low side
Attached scheme depicts 10 top scoring compounds with the highest CLint values in the set. LogD range: 2.8-5.2 CLint range: 332-668 μL/min/mg
Interesting observation: ketoxime (top right) has high CLint value (435 μL/min/mg ) at logD 5.2, which is twice as much as we we would usually expect when calclulating lipophilicity during analog planning.
For what it's worth the azoles (of which our fenarimol derivatives can be thought of as a sister class to) have variable half lives but as a general rule it's said to be around 20 hours on the high side and 16 on the low side
@MFernflower I will upload a separate graph on t1/2 soon. Looking at half-life data preliminary, I can tell that the top scoring compounds have t1/2 values usually not exceeding 5 min
@dmitrij176 There are online metabolic prediction tools - One is called xenosite and it's pretty good and free - maybe worth messing about with https://swami.wustl.edu/xenosite
@dmitrij176 I ran dm7 past the tool and it flagged the pyridine motif as being prone to enzymatic oxidation:
If we're aiming for an orally available pill, then metabolic clearance is relevant. i.e. human microsomal data.
Can we dig up the data for CLint or t1/2 that exist on any MycetOS compounds from the original Epichem publications?
Can we then plot those data vs logD?
It'd be useful to know whether low logD also helps us via better in vitro pharmacokinetic data.
I say "we" - I mean anyone interested in the science of drug development here who is not me... Perhaps @fantasy121 or @dmitrij176 would be interested in kicking this off? Ultimately we'd need a page on the wiki on this aspect of the series, and we should probably address it in the forthcoming paper.