It is common for medical researchers, medical professionals and people looking at data on the Internet, to misunderstand receptor affinity data, in relation to both its accuracy and its functional consequences. The extent to which such data is misrepresented in some scientific papers makes it difficult to be certain whether lack of understanding or disingenuousness is at the root of the problem. My web site essay and my published paper (1) on the receptor data in relation to the drug mirtazapine illustrates this point, but similar observations are relevant for the data on many other drugs such as fluoxetine, quetiapine etc.

The first major issue is that it is necessary to compare apples with apples, and not apples with oranges or bananas. What that means in this field of research is using only human cloned receptor (HCR) data. Receptors vary between species and measurements made in test tubes (‘in vitro’) using tissue from rats or other species cannot be compared to what happens at a similar receptor in humans. Some papers mix measurements made in different tissues and in different species, all in the same table of data, without making it clear to the reader what is being done (that is what Organon did with the mirtazapine data, and see also (2)). Lilly did the same with Prozac (fluoxetine) as analysed by Stanford (3). It seems to me that can only be explained by incompetence or dishonesty.

Since the relevant receptors were not produced in their human cloned form until the mid-1990s any paper before then is almost certain to be a mixture of measurements made in different species (with the exception of some work done with human brain post-mortem tissue cf. early papers by Elliott Richelson (4, 5)).

For our present purposes non-HCR values can be dismissed as being insufficiently accurate and not comparable to subsequent data derived from human cloned receptors (HCR) assays. Also note that IC50 data are often of little use because the values are so dependent on assay conditions such as substrate concentration they cannot usually be meaningfully compared.

Affinity is only part of the story and one cannot leap reliably from affinity to potency to efficacy.

My review of the tricyclic antidepressants was the first paper to assemble the HCR data on these drugs and the tables in that paper are thus more precisely comparable for side effects and benefits between drugs than any previous review. Furthermore, that paper also used other types of data to establish a relationship between receptor affinity and magnitude of side-effects, toxic effects and beneficial effects. That is quite a complex process and anyone interested should read that paper carefully. My review in Biological Psychiatry (6) used that approach to look at serotonin toxicity with various different classes of drugs. Indeed it was the understanding resulting from those considerations that led me to make the prediction that methylene blue must be a potent MAOI. That prediction was correct and has been a life-saving revelation for many patients. MB is being used for all sorts of purposes, indeed nearly 10% of Americans are on SSRIs, so many people are at risk of the serious interaction between MB and SRIs.

However, as noted in that BJP paper, it must be remembered that even the best laboratories will get significantly different results with apparently the same techniques. My estimate of the variability in human cloned receptor data measurements made in different laboratories was that they varied by a factor of 10. It is thus clear that any papers that discuss comparisons between drugs and do not take that variability into account are not realistic.

Another characteristic of such papers (e.g. (2, 7)) is that many of them quote results to spurious levels of accuracy. In other words measurements which are only accurate to one decimal place are given to 3 decimal places. The next factor to take into account is that for us to be confident in scientific results they need to be replicated by independent workers. It needs to be shown that the results are within a reasonable degree of agreement. Many of the values being touted for drugs have not been independently replicated (e.g. norquetiapine, aka N-desalkylquetiapine, as an NRI). If we remind ourselves of the approximately tenfold variation in values produced by different laboratories that will help to maintain some degree of objectivity and realism. This variation makes a nonsense of most of the claims about atypical antipsychotics (see separate commentary).

In summary therefore, human cloned receptor (HCR) data is valuable in helping us to understand the potency of drugs for particular receptors but it needs to be interpreted with the above caveats in mind. Even then, we cannot always be sure whether different drugs achieve the same levels in the human brain under every day conditions and there are various other variables to be taken into account. Anyone without in-depth understanding of therapeutics and pharmacology is unlikely to be able to understand the various other factors and variables that are relevant which include the notions of “agonist directed trafficking”, intrinsic activity and binding tightness. There is much we still do not understand about nerve cells and neuro-transmission: indeed the reality and consequences of agonist directed trafficking have only just begun to be recognised.

One of the things that my research has been concerned with, and has clarified, is the relationship between the in vitro (test tube) receptor activities measured in experimental work and the actual effect of drugs in humans. This is a vital consideration because the potency of many of these drugs varies by several orders of magnitude [see table of Receptor Affinity data below]. So, the key question is how potent do these drugs have to be before they have significant actions?

Note that the table below contains a range of values for the receptor affinity measurement. This needs to be taken account of when comparisons are made and it is frequently not valid to try to make precise comparisons. The data has to be interpreted in the light of other data and properties of the drug so that a best guess can be made as to exactly what the data indicates. When reading different papers care needs to be taken to note whether the data relating to human cloned receptor assays is being presented, or some other technique. HCR data is probably the benchmark and all data herein is HCR data unless otherwise indicated.



Table Human cloned receptor (HCR) data relating to ‘dual action’

Reuptake inhibitor affinities Ki nM (modified PKG March 2010)






SSRIs (for comparison)




~ 1:1000



19 - 102

2.8 - 36

















20 - 142

1.3 - 20




0.63 – 8.6

22 -180













No HCR data

No HCR data







Table Legend TYR30


+++ potent effect ++ moderate + weak 0 no significant effect.





1.                  Gillman, PK, A systematic review of the serotonergic effects of mirtazapine: implications for its dual action status. Hum Psychopharmacol, 2006. 21: p. 117-25.

2.                  Gross, G, Xin, X, and Gastpar, M, Trimipramine: pharmacological reevaluation and comparison with clozapine. Neuropharmacology, 1991. 30(11): p. 1159-66.

3.                  Stanford, SC, Prozac: panacea or puzzle? Trends Pharmacol Sci, 1996. 17(4): p. 150-4.

4.                  Richelson, E, Neuroleptic Affinities for Human Brain Receptors and Their Use in Predicting Adverse Effects. J Clin Psychiatry, 1984. 45: p. 0.

5.                  Richelson, E, The Newer Antidepressants:  Structures, Pharmacokinetics, Pharmacodynamics, and proposed Mechanisms of Action. Psychopharmacol. Bull., 1984. 20(2): p. 213-223.

6.                  Gillman, PK, A review of serotonin toxicity data: implications for the mechanisms of antidepressant drug action. Biol Psychiatry, 2006. 59(11): p. 1046-51.

7.                  Meltzer, HY and Massey, BW, The role of serotonin receptors in the action of atypical antipsychotic drugs. Curr Opin Pharmacol, 2011. 11(1): p. 59-67.