How to Design a Drug: Insights into Drug Discovery

Posted on Updated on


Today we are joined by Sarah Dykstra, PhD, who worked in the early drug discovery group at Amgen. Many of our postdocs and PhD students will end up in R&D, so its a great opportunity to learn about your potential future career.

Big thank you to Thermo Fisher Scientific for sponsoring this seminar! Remember that if you need any support from Thermo Fisher, feel free to reach out to our Account Manager, Adam!

Here are some insights from the seminar on the drug discovery pipeline:
(A copy of the slides can be found at the end of the post!).

Sarah got her Bachelor of Science from Stony Brook University and then went on to be a research associate in Amgen’s Cambridge Hematology Oncology Group… and then got bored and got her PhD from Brandeis University.

More on the industry job…

  • She performed screens of small molecules and investigated candidates for potential therapeutic intervention.
    • Identified targets from the literature and validated that the data was reproducible (as much of the published data is not!).

So, how do you make a drug?

First, you need to determine what you are screening for and what your downstream reporter might be.

Steps in Drug Discovery:

  1. Discovery Research
  2. Preclinical Research
  3. Clinical Research
  4. Regulatory & Manufacturing
  5. Launch – sale of product.

Only about 1 in 5,000 molecules make it through discovery to launch and this entire process takes over a decade!

What do you do in discovery research?

You think about what candidate genes would be amendable to therapeutic intervention by reading the literature and attending conferences and meetings.

 You need to find a good target, one that has minimal side effects.

The other major function is to design cell based assays to validate therapeutic targets, then screen small molecules to determine several initial “hit” compounds.

(A hit is something that is potent at a micromolar range whereas a lead is a modified hit that has potency at the nanomolar range, this is important to minimize off target effects).

How do you decide what target to work on?

It is generally a joint decision based on the team and the literature.

Goals of discovery research:

  • Identify a candidate to interrogate
  • Design a screen (and counter screen)
  • Find a series of hit compounds:
    • Refine specificity and potency via modification of chemical structures (to find a number of leads)
  • Determine a lead compound to bring through to pre-clinical development.

Difference between pharma and biotech:
Pharma compounds used to be “chemical” compounds and had libraries of tens of thousands of compounds in libraries, whereas biotech companies are from the era of recombinant DNA and used biologics (protein based therapeutics). Now, both compounds could have millions of compounds that they routinely screen.

Preclinical Research:

How your drug works in vivo!
Determining the optimal in vivo dosing, safety profile, PharmoKinetics (the life span of the drug), PharmoDynamics (how the body metabolizes the drug and how bio-available it is) and ADME (absorption, distribution, metabolism and excretion of the drug – much of this influences the side effects).

Design stable API formulations and good manufacturing practices (GMP) to develop analytical methods

… for incorporation into an IND filing.


Goals for preclinical research:

  • Determine ADME, PK, PD, toxicity
  • Finalize a lead compound and scale up manufacturing
  • Determine a dosing schedule for first-in-human trials
  • Develop a clinical plan and preparing a new drug product
  • Prepare for IND filing

Question: How many animals are required?
A: Several hundred and it generally starts with rodents and rabbits and ends with higher mammals such as dogs and pigs. Monkeys are being phased out for a number of reasons.

Clinical Research

– Designed to determine safety and dosage in humans, efficacy and side effects in increasingly large human populations. This is required to account for polymorphisms and other genetic modifiers than can influence drug bio availability and metabolism.
– A clinical trial is based on data obtained from pre-clinical research on dosage, PK/PD and toxicity which requires FDA approval of your NDA.

Phase 1: 20 – 100 healthy volunteers to evaluate safety and dosing.

Phase 2: More patients to assess drug effectiveness and further evaluate safety.

Phase 3: Largest number of patients involved, further studies on safety and efficacy in larger populations.

Goals of clinical research:

  • Obtain safety, efficacy, dosage and adverse effect data
  • Use this data to convince the FDA that there is evidence from preclinical and clinical research that a drug is safe and effective for its intended use (NDA filing).
  • Finalize lead compound and manufacturing.


How many drugs are pulled off the market after Ph4 trials?
In the case of cancer drugs, these are rarely pulled due to the terminal status of the patient.


The riskiest step is within the discovery and preclinical steps whereas the typical failure point is around phase 2 – during clinical trials.

Making a drug is RISKY business, costing $1-2 billion dollars and close to a decade (if not more) with a low success rate.

Rational drug design is a big, relatively new area in which you design the drug to fit the site, rather then screen drugs that might incidentally fit your target.

What Sarah learnt from working about industry:

  • Team work and soft skills are highly sought after (and required!)
  • You won’t be abandoning science, but you do need to accept that priorities change rapidly.
  • Success is carefully defined and those who exceed expectations will be advanced and well rewarded.
  • Not all companies are run similarly, expect cultural and management differences, a start up does not run like a large company.
  • Industry positions are highly competitive, expect to work with the best
  • You will constantly have to learn, network and delivery to succeed.
  • The most successful industry scientists have excellent time management and communication skills and are highly productive.

This is talk is part of a series! The next seminar is a case study on science and strategy behind the development of two commercial drugs: epoetin alfa (Epogen), a biologic from Amgen and Gleevec (Imatinib), a small molecule from Novartis.

Click here to download the slides from the talk! – Drug Design Talk_1 Sarah (pdf).