-The discovery and development of new drugs is in a stage of rapid acceleration, due to the confluence of four factors: easy access to big data through internet, an explosion of biological structural information, advanced modeling, and extremely cheap computer power through clouds. In one sweep, the course takes the student from the inception of a target and a potential drug candidate, to clinical trials. The student will learn the following concepts: (1) the organization of pharmaceutical molecular research in the discovery pipeline workflow paradigm, (2) the sourcing of (structural and property ) data through webportals, (3) important molecular property calculations, such as charge and pK, and calculation of derived properties such as solubility and logP, (4) the principle of screening ultra-large databases, and relation with chemical informatics concepts such as similarity and molecular fingerprinting, (5) docking, and the relation to fragment based drug discovery and high-throughput screening, and (6) correlation models for toxicity and relevance for clinical trials.
In the course, we will make frequent reference to real business cases, from small companies from the local Leiden BioScience Park, and international pharma. In particular, we will use the development of new kinase inhibitors for cancer therapies as testbed. In short, the returning theme is: why are new cancer medicines so horribly expensive? And what can we do to improve on that? Some organizations, such as Cloud Pharmaceuticals, claim the solution is the rational design of drugs, solely using inexpensive cloud computing, where the cloud computing foregoes expensive testing in pre-clinical trials. Other solutions could be in citizen science, where lay-persons offer free PC-time for folding proteins (folding@home organization). Yet other approaches circumvent traditional thermodynamics and quantumchemistry completely, and use Artificial Intelligence through big-data and machine learning (Google AlphaGo and IBM’s Watson). Perhaps yet another solution is to forget about the idea of ‘design’ altogether, and simply conduct massive trial-and-error experiments through the extensive use of advanced robotics (see https://en.wikipedia.org/wiki/High-throughput_screening). Most likely, of course, the solution will be a combination of all these.
The course is in the form of 12 lessons, of each 2 times 45 minute lectures. In addition, there will be three blocks of practical courses, of each 2 hours.
At the end of the course, the student will be familiar with the way pharmaceutical research is organized, and will have a good overview of the most common calculation methodologies. The student will be able to digest presentations from recent conferences, some of which will be used in the course as supporting materials.
No prior knowledge in quantum chemistry or thermodynamics is required, but the student does need to have a strong affinity with programming and mathematics. Methodologies such as the charge equilibrium method and Linear Free Energy relations will be introduced in the course. In addition, we will discuss extensively common numerical mathematical operations such as non-linear optimization, and solving large sets of linear equations.
A prerequisite is a basic understanding of the python scripting language, version 2.7, that will need to have been installed on the student’s personal computer prior to the course. Additional packages will be provided at the course. The student is strongly advised to have achieved good understanding of python through self-study or additional courses.
The student should be aware that the focus is on the mathematical background of the various methods.
Mode of Instruction
Lectures and hands-on workshops.
The teacher will provide a reader (collection of papers), a powerpoints and examples of exam questions. Students interested in more background are advised to acquire ‘Molecular Modelling, principles and applications’ from Andrew leach, second edition (Pearson/Prentice hall, 2001)
Exam is 3 hours closed-book, with one question per lesson. Example questions will be distributed during the course.
Some of the teaching material is based in the course ‘Molecular Modeling’, given in previous years by the same teacher. Presence at the lectures and workshop is obligatory.