QED (quantitative estimation of drug-likeness) is a score which helps you to predict if a small molecule is desirable as an orally absorbed drug, described in the Quantifying the chemical beauty of drugs article.

Its calculation is based on well defined chemical properties such the number of hydrogen bond donors and acceptors, aromatic rings, rotable bonds, octanol-water partition coefficient etc...

All of this properties can be derived from the structure of the molecule. One can use for example the QED module of RDKit

I develop a compound database for drug discovery. Some compounds are salts where one ion is the organic molecule and other is common inorganic ion such as $\ce{Na+}$, $\ce{Ca2+}$, $\ce{Cl-}$, etc...

When I load the salts into the database I derive their non-salt form, just the organic molecule.

My guess is that I should only calculate the QED score and its associated parameters, or any chemical parameters for ADME for the organic molecule, but not for the original salt compound.

My logic is, when we apply a drug as a salt orally it will dissociate into ions, so the organic bio-active molecules will be separated from the inorganic ions. Of course the inorganic ions will be present in the gastrointestinal fluid, but in such a small quantity compared to the other inorganic ions (often the same type) that they will have no significant effect.


1 Answer 1


I found the answer when I was reading about InChi in more detail.

For example, to a bioscientist “glutamic acid” and “glutamate” are the same thing, but to a computational chemist the loss of a proton, or its variable site of attachment is critical.

So the guess in the question is OK in a bioscientist perspective, but not for the computational chemist perspective, and since this is biology oriented computational chemistry question both the ionic and neutralized form must be considered.

To test this I calculated the Quantitative Estimation of Druglikeness (QED) for in a simple case for , sodium acetate, acetate (just the organic anion) and acetaldehyde. Their mean QED:

sodium acetate: 0.2777

acetate: 0.3498

acetaldehyde: 0.4298

The ionic forms (sodium acetate, and just the acetate anion) are less likely drugs since ions, since ions has an extremely low chance to pass biological membranes passively. The sodium acetate has lower likelihood then acetate, because it contains 2 ions and the QED algorithm has now way to tell which should be the drug and assumes that both ions must pass the membranes which is a great penalty.

In conclusions the QED parameters should be calculated both form the neutralized and organic ionic forms. The in-vivo relevance of the two forms will heavily depend on the chemical environment, since they will be in a dynamic equilibrium, which mostly depends on the PH of the micro environment. The QED parameters should not be calculated from the salt form, since we are not interested in the QED of the inorganic ion, further many more inorganic ions will be present next to that.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .