Usually small molecules should be in excess, unless you assay is very sensitive to a change in a small fraction of the protein targets. However most of the time this is not an issue because your proteins would be in the low uM or high nM range while the drugs would be in uM to mM. I would try as much chemical as possible -- as long as their solubility allows. The chemical solubility might be a real challenge. Some drugs may form tiny aggregates at uM concentrations, and inhibit biological reactions nonspecifically. You need to be careful with specificity controls.

The reasonable molar ratio of molecule to inhibit any protein may be defined by determining the IC50 values. first the compounds can be assayed at different micro molar concentrations and the plot generated Conc. Vs Inhibition will give the IC50 value of that compound which will give the reasonable molar Conc of inhibition

To add to the points raised by Zhengjian, I'll encourage the original poster to provide some further information about both his assay and expectations for potential compound affinities to help focus this discussion. I'll also encourage him to keep in mind issues of absolute concentration, not just molar ratios. Zhengjian brings up one of these - solubility - but please remember that absolute concentrations are essential to consider for protein saturation as well (e.g. 10:1 ligand:protein ratios will lead to radically different loading of Kd=1nM, 1uM, 1mM compounds to the low uM protein concentrations used for many in vitro assays).

If an in vitro biochemical assay is being used, as I'm guessing from the original post, I'd agree with Zhengjian's tips; if this is intended for use in cells, the answers will vary significantly. Careful secondary screens and controls are essential in either case.

The best way would be to perform isothermal titration calorimetry. With ITC, you could assess stoechiometry easily.

"Everything interact with everything else" Thomas Fridman.

A complex refers to a system with many agents (independent agents... / dependent agents...) - Topological connectivity in MA-CAS - Anticipated information of systems. The information is embedded in the multi-topological connectivity of the "many agents" in a microenvironment. Form this multi-topological connectivity emerge a property. An emergent property is not found just by looking at the many different compounds separately. It's like a puzzle where all the different pieces are only a fragment of the completed image.

You should work with "sets" of 3 to 5 small molecules at a time, to get the best output... and not!! one at a time? "multi-topological connectivity emerge a property"!!!

CE0702

If your final goal is to develop a drug, molar ratio really is not important. Find the IC50 from two-fold serial dilutions starting from a high of around 50 uM if possible and titering down to picomolar in your screening assay. As a rule of thumb we discard compounds with IC50 greater than 5 uM, as they will not make good drugs (may still be good as a research tool, though). For the good hits you should immediately go to cytotoxicity assays on e.g. NIH/3T3 (or your preferred cell system). Any compound where IC50 for cell growth is not at least twofold higher than IC50 in your protein-protein assay will not make a good drug and can be eliminated. Then try your compound in a sub-toxic concentration range in a cellular model, if you have one. These are just rules of thumb given the limited information you have provided, but this strategy is only a week's work to pare down your short list of compounds to a few that might actually be useful in drug development. You don't need to do fancy physical methods until you know you have something useful and can spend your time and money on just those useful compounds. If your whole modus operandi is describing the physical nature of the protein-protein interaction that is a whole different ballgame.

Thank you everyone for your replies. Each response has provided valuable insights and also brought up new questions.

1) My assay is a two step enzymatic assay which requires addition of a cofactor Y.

E1+ Y activates E2

Activity of E2 is determined by cleavage of colorimetric substrate

It is sensitive to relatively low concentrations of cofactor (pM range) and addition of a protein inhibitor (low nM range) completely and specifically inhibits cofactor activity; though I have not yet determined the kinetics (Kd or IC50) of the protein inhibitor. I believe that I can enhance the sensitivity of this assay once I determine the IC50 value.

2) As for solubility I can start with estimates such as xlogp3 found in pubchem. I understand the the benefits of an xlogp3 estimation are minimal and I may have to test solubility of individual compounds being screened. How might you suggest I tackle the problem of solubility? This is unfamiliar territory. For instance, how would I know if there are aggregates?

3) Using small "sets" of these compounds, determining the IC50 of inhibitory compounds, or using ITC to determine stoichiometry is all work that will be important after addressing the more acute issues of the current screening process and I appreciate the necessity of these experiments.

Norbert you address a very important point: what is the end goal of this assay(s)? Although the eventual goal is to screen for viable drug candidates there are some intermediate steps that I believe can refine my search. I plan to do more virtual drug screening using pharmacophores generated from those compounds that are biologically active. This means that anything which inhibits this protein-protein binding is (at least initially) of interest. However, understanding what properties must coalesce to make a good drug will be important in selecting candidates after this refinement and I am grateful that you brought this to my attention.

I also think the absolute concentration of small compounds is of more interest.

Kd=[P][L]/[PL]; when [PL]=[P], Kd equals the free [L].

So, when half of the proteins are in the bond form, the absolute concentration of your compound is {1/2[totalprotein] + Kd}; say if you have 1mM protein and the Kd is about 1mM, you need 1.5mM compound to half saturate the protein.

So the absolute concentration depends on the amount of protein and Kd. since most kd in the uM-mM range, you can well estimate the amount of compound to half saturation. The work concentration can be 3-5(?) folds excess of this concentration.

To evaluate the solubility of organic compounds we have used 1D 1H-NMR spectra. If a solution of your compound contains precipitate anduou want to know how much remains in solution you can add a known amount of a reference molecule and from the relative intensity of known signalsin the reference and in your compound you can calculate the concentration of your molecule in solution. In water or water/DMSO solutions a useful reference molecule is tryptophan because you can easily measure the concentration of the stock by UV-280nm, and it provides signals from its aromatic protons that you safely integrate if your molecule does not have aromatic hydrogens. For your compound signals you can measure the intensity of all the methyls, for example, and normalize for the number of protons (in the compound methyls and in the tryptophan aromatic group). Another possibility is to use the signal intensity of the DSS or TSP (the NMR reference for chemical shifts at 0.00 ppm in water solutions) if you added a known amount from a stock (determined by gravimetry or by the tryptophan method).