feasibility in compound synthesis. Library design and how easy you can access to it.

The answer really depends on the pathology and the drug target. Here are some examples, particularly from the CNS field.

In the case of highly-investigated diseases for which there are already numerous treatments (e.g. schizophrenia, depression), an original mechanisms of action and therapeutic superiority are essential. Hence, identifying yet another D2 receptor antagonist or serotonin reuptake blocker, respectively, is insufficient, even though the targets are well validated and the new drugs may be more selective or potent than existing ones.

In the case of orphan disorders (e.g. Fragile X, Huntingdon's disease), any demonstration of possible therapeutic efficacy is valuable, regardless of mechanism of action. In fact, the latter need not even be known, provided there is activity in a relevant pharmacological model.

Similar considerations apply to drug safety / tolerance. If the medical need is high then good tolerance is less of a priority and smaller therapeutic margins will be acceptable. The obvious example of this (outside CNS) is in oncology, where cytotoxic drugs are poorly tolerated but widely used.

A commercial consideration concerns the availability of cheap, generic products that may address a substantial proportion of the medical need in specific disease areas. Risperidone is an example of a cheap and widely available antipsychotic which is efficacious in many patients. Not surprisingly, many companies are targeting other symptoms of schizophrenia, such as cognitive deficits, instead of psychosis.

So to start answering your question, you need to critically consider the disease area you are interested in, the level of medical need and the availability and efficacy of existing drug treatments.

Hello Fidele,

I've heard of labs for which the most important property is NOT linked to the disease, to the biological targets, not even to the biological effects induced by the drug over the considered targets.

For these labs, the very most important point (and by far) is the toxicity of the tested molecule.

Because you could have the most potent drug ever, the most selective drug ever, the most 'what-you-want' drug ever... if it's too toxic, you won't do anything with your molecule and it will be nothing more than just an active molecule, maybe useful for in vitro test, at best, but nothing more.

Note that getting a new and active molecule for in vitro test is a good thing, but it's not what you want here.

So it will NOT be an authorized drug by the FDA or any other organisation.

That's why some labs check for toxicity even before checking activity or selectivity.

Other labs check for other properties, such as hydrosolubility for example.

Because, here again, even if your molecule is active, you won't do much with it if it's not enough hydrophilic.

Of course, you could use nanoparticles to help improving solubility, but it's more complex.

I agree with everyone, the goal is most important when making a regulatory strategy, but also the money that are available for the project and the time of fulfillment. I suggest an article that may help clarify the issue.

How To Develop A Successful Regulatory Strategy. Outsourced Pharma, June 19, 2013. http://www.clinicalleader.com/doc/how-to-develop-a-successful-regulatory-strategy-0001

I think that the most important things are:

- the characteristics of the molecule or of the compound

- the mechanism of action of the product, its uses or its new uses (if it isn't a new solution)

- its effectiveness in achieving the goal ((effectiveness / efficiency)

- its toxicity (weighted to the benefits, referring to the types tested)

- Cost / benefit report.

Hoping it will be useful,

Luisa

Dear All,

I am very impressed by your contributions. I think that each step along the drug discovery chain is important, and nothing to be neglected.

However, target validation seems to be the most crucial stage.

Once more thank you all

Fidele

Are the availability of virtual compounds hits ever considered in virtual screening campaigns? This too could be very important.....

do you mean market availability? If so, ZINC db could be helpful.

To the list above I would also add prediction of off-site targets and ADR in early stages of development

Yes, it is usually preferable to carry out virtual screening on a database of commercially available compounds, if not you end up with virtual hits which only remain on paper.

It is also important to consider if the compounds are in sufficient quantities to be screened.

Another important question is how well in vivo data from rodent models translate well to humans

ADME and toxicity prediction steps might be vital for the success of the drug in addition to the output of clinical trials.

Without explicitly stating it, everyone (including the OP) is talking about a program designed to identify a drug that targets a specific protein. In this case target validation is crucial, and often rather challenging. However, if you develop a phenotypic screen rather than an in vitro protein assay, then the hits are "self-validating". The crux of this approach on the other hand is developing and validating a robust screening system.

Yeah, I agree with James perfectly. Designing a screening program is of utmost importance

Some groups do not consider DMPK outcomes to be of utmost nowadays with the emergence of nano-particle delivery systems. What do you guys think?

A lot of good advice coming through the forum!  I would suggest that there is value in getting clarity on whether you are carrying out a drug discovery project or a target validation or disease alignment project.  If drug discovery, then your patient population, what is required by patients, physicians and paying organisations in terms of efficacy, side-effect liability and cost should be clear.  This allows the screening strategy to be formulated and clear decision point milestones for your target or chemical series to be be set.  Essentially you know the destination for your journey and need to make sure that you stay on the right path to deliver.

For target validation or disease alignment, then good quality, highly validated and well characterised tool molecules are needed - these may need to be more selective than drugs for the alignment of the biology to the disease state to be unequivocal.  There was an excellent review on the pitfalls of poor chemical tools recently (doi:10.1038/nchembio.1867) which is well worth a read.  In this case you dont know the destination so need to be sure of the soundness of the path before you move forward!

I believe both approaches can and have delivered drugs.  The number of drugs that end up being used for diseases that they were not initially designed for suggests we are still doing target validation a long way into clinical trials.

Good luck with your drug discovery efforts - do not be discouraged by projects that do not succeed - as your correctly say, the barriers to new drugs are high.

Does coverage of functionality space in designing compound screening libraries to represent a reasonable SAR play a role for DMPK?