Challenges in drug design and how CADD can help

In a classical approach, you could work with a team on a new drug basing on general chemical knowledge and test your ideas using trial and error methods. But even the greatest minds can't really imagine, and what is more, test all the possibilities. There definitely should be some pre-elimination of the less promising paths. But which ones are the dead ends? And how to estimate the costs and duration of the whole process when the percentage of failed trials is both high and unpredictable? It is exactly why you need CADD in modern medicine. It won't make up for the whole clinical process, but it can reduce the project's costs even up to 50%.

CADD enables scientists to:

  • model the structures of proteins and compounds that are not yet known,
  • create virtual compounds libraries,
  • model the compound interactions and reduce significantly (very significantly) the number of ligands to be tested,
  • do all kinds of molecular docking and virtual screening calculations,
  • determine many properties of a ligand to predict its toxicity or other important features.

No wonder CADD covers most stages of drug development: from target identification to validation, from lead discovery to preclinical studies. It shortens the process and saves money, which is always a huge benefit in the long run. And drug design is a long run - from a few to a dozen or so years of research.

What is worth knowing, computer-aided drug design is not exactly a novelty. In a broad sense of "computer-aided", each time the computer technologies are of help in drug design, it is CADD. Therefore instead of talking about a specific technology, it is better to think of it as a method of using a computer to do some large-scale calculations and modeling necessary for drug design. CADD has been used for HIV drugs approved already in 1996, which gives it quite a history. Today, with a high pace of discovery and demand for new therapeutic agents, computer-aided drug design only gained importance.

Various applications of CADD in modern medicine

CADD tools are applied in almost every stage of drug research and development, thoroughly changing the strategy for drug discovery.

Originally, it was mostly used in the core lead discovery and optimization phase. Nowadays, it expands in both directions of the drug go-to-market flow.

Computer-aided tech assists the researchers in i.a. such processes as:

  • screening promising compounds for target identification and validation through reverse docking and bioinformatics,
  • predicting target druggability and compound 3D structure before further experiments,
  • foreseeing ADMET (Absorption, Distribution, Metabolism, Elimination, Toxicity) properties and conducting physiologically based pharmacokinetic simulations to model the preclinical tests accurately.

Let us give you an example of the new possibilities provided by CADD. Avastin was originally developed for metastatic colon cancer and non-small cell lung cancer. Thanks to additional research - molecular docking - it has now been approved for metastatic breast cancer. We wouldn't have known this application if it wasn't for CADD. Additionally, some drugs for liver cancer are currently in the final stage of clinical trials - also thanks to the CADD software.

To learn more about the new technologies in CADD, go to this research.

There are numerous applications of CADD in modern medicine. We may even dare to say they are endless. But the crucial part of it is who handles it and how they do it.

Is CADD never wrong? Why do you need people who know how the "black box" works

The current trend of uncritical use of machine learning and artificial intelligence may result in unfortunate misinterpretations of data. The field of computer-aided drug design (CADD) is not an exception. On the one hand, more and more scientists are becoming aware of the benefits of available shared data and are beginning to derive predictive models before going to the lab. On the other hand, easy accessibility of in silico tools comes at the risk of using them as "black boxes" without sufficient expert knowledge, leading to widespread misconceptions.

This is why a team of dedicated professionals, who not only understand all ins and outs but also design the so-called black box, is necessary to use CADD for your benefit.

Deep Matter & Code Poets - CADD in action

While working on the ICSYNTH project with Deep Matter, we take part in designing and optimizing the retrosynthesis software. It enables the users to model synthetic pathways to the target molecule in order to select a good option for the research.

The software makes it possible to project a pathway - a multistep synthesis tree that meets the user's requirements. The reactions are based on published materials, and the list of substrates comes with the real-time availability assessment. The same target molecule can be worked on by a user group simultaneously, which helps increase efficiency and save money.

This project is like a tailor-made Code Poets playground. We are both scientists and developers, which is why this kind of challenge comes with great joy and a sense of fulfillment. We can't wait for the to-be CADD-related ventures. Apparently and luckily, there is still a great need for human brains in computer-aided drug design.