Indeed - there's no sure answer to your question. In general, work from the following:
Do you have compounds of the same class? Test from at least 10X a known compound's IC50 on downwards.
Do you have other compounds of ANY class? Try maybe 100X that IC50.
Do you have nothing at all to work with? Then think about the maximum dose that would be meaningful to you - many people say 50 uM or so. Test from that on down, in 2-fold dilutions. 3-fold if you want a really, really broad range.
But there's really no way to guess at the dose of an unknown compound for an unknown target - and even if we knew the structure and the target, it'd still just be a guess.
Hi, Generally, less than 10 micromolar (uM) concentration of a drug in plasma is accepted because higher than 10 uM of a drug inhibit necessary enzymes. The most of drugs on the market have less than 10 micromolar therapeutic plasma concentrations. Therefore, I suggest you setup an experiment includes 10 uM, higher and lower doses for initial screen. We generally start with 40-50 uM of a new molecule as a highest dose and decrease to 1 uM. But you can also use 10 fold rule which you test concentrations from 0.1 uM to 10.000 (0.1, 1, 10, 100, 1000, 10000 uM).
If there is no any information about the possible activity of your compound in cancer cells, I would make a stock solution of 20 mM and then would make serial 10-fold dilutions in a medium starting from 100 uM (100 uM, 10 uM, 1 uM, 0.1 uM, 0.01 uM, 0.001 uM, etc.).
I would suggest to use 96 well plate and start with 100uM, then do 2 times serial dilution of 18 different concentrations with concentration 0 to serve as a control.
let me suggest use 96 well plate and start with 50 and 100uM, then do 3 times serial dilution of 18 different concentrations with concentration 0 to serve as a control.
For the maximum concentration usually I perform a preliminary solubility test. I try to dissolve tested compounds in DMSO ( as usually they are not very soluble in water). Then take 990 ul of PBS (or medium) and add 10 ul of prepared stock solution - check if compound is not precipitating. Otherwise, it could precipitate also in a medium, and you will not know the real concentration in it. If it is precipitating, I make more diluted solution in PBS and check again. In such way I find the maximum concentration of compound which could be prepared in aqueous medium. I would not use that concentration of compounds which is too high and compound precipitates during dilution or in mins.