Bridging Course A1 Ā· NUS Department of Biochemistry

Basic Medicinal Chemistry

The chemistry a health science graduate needs before computer-aided drug design

Twenty-three modules in five parts, from reading a skeletal formula to triaging a hit list. Everything is anchored to a marketed drug or a real complex. By the end you will be able to open an unfamiliar structure and say what it is made of, what it will do in water at pH 7.4, and what a computational method is likely to get wrong about it.

The first unit of the Design and Modalities strand, MScPHMxAI

The essential modules unlock in order: to open the next one, finish the one you are on — answer the pre-test and the post-test, rate the module, and write one thing you learnt. Modules marked extension or optional never block anything; take them when the topic comes up in your own work.

Part 1. Fundamental organic chemistry for drug molecules

What the molecule is made of, and the rules that constrain it.

1

Reading a structure

Skeletal formulae are a convention, not a picture. Learn to read one, and to draw ten marketed drugs from their names.

2

Atoms, bonds, valence and charge

Valence, hybridisation, lone pairs and formal charge arithmetic, done by hand until it is automatic.

3

Functional groups and what they do

The twenty-five groups that account for almost every marketed drug, and the four properties you need to know about each.

4

Aromaticity and heteroaromatics

Huckel's rule, pyridine against pyrrole, and the ring systems that recur across unrelated targets.

5

Reactive and undesirable functionalityExtension

Electrophiles, Michael acceptors and structural alerts — and why the same chemistry is a warhead in one molecule and an artefact in another.

Part 2. Conformation and stereochemistry

What shape the molecule can adopt, and at what cost.

6

Rotatable bonds and ring conformations

What can turn, what it costs to turn it, and why two programmes report different rotatable bond counts for one molecule.

7

Chirality, E/Z and atropisomerism

Find the stereocentres, assign R/S by hand, and meet the eutomer, the distomer and the atropisomer.

8

Energy landscapes and the bioactive conformation

The conformer that binds is rarely the one at the bottom of the well, and the difference is paid for out of the binding energy.

Part 3. Physicochemical properties

The numbers that decide whether the molecule ever reaches its target.

9

Lipophilicity: logP and logD

The single most consequential number on a medicinal chemistry project, and the trap it sets for anyone optimising potency alone.

10

Ionisation: pKa at physiological pH

Henderson-Hasselbalch by hand, microspecies, zwitterions, and the fact that the neutral form you drew is usually not the species present.

11

Tautomers: which structure did you model?Extension

A shift of one proton turns a donor into an acceptor without changing the formula, the name or the molecular weight.

12

Molecular weight, TPSA, donors and acceptors

The descriptors every filter is built from, and why the same molecule has more than one correct hydrogen bond donor count.

13

Aqueous solubility and its structural determinantsOptional

Escaping the crystal and being accommodated by water are two separate problems, and flatness ruins both.

Part 4. Molecular recognition: the interaction toolkit

The physics a scoring function is trying to approximate. The heart of the course.

14

Hydrogen bonds and the desolvation cost

Both partners are bonded to water before they meet. What a hydrogen bond is worth is what is left after paying to strip it.

15

Van der Waals contacts and shape complementarity

Many small contacts, summing to more than any single polar interaction. Why potency tracks buried surface area.

16

Electrostatics and salt bridges

The same charge pair is worth wildly different amounts in bulk water and in a buried pocket. The dielectric problem, and what it does to scoring.

17

The hydrophobic effect and ordered water

Binding is driven by water leaving, not by oil attracting oil. A binding site is never empty before the ligand arrives.

18

Pi interactions, halogen and chalcogen bonding

Stacking geometries, cation-pi, and the sigma hole that most force fields cannot represent at all.

19

From interaction to number: free energy and Kd

Convert a dissociation constant into a free energy, and learn exactly what a docking score is not comparable to.

Part 5. Drug-likeness, filters and efficiency metrics

How to triage a hit list, and how to defend the compounds you rejected.

20

Rule of Five, Veber, Ghose, EganExtension

What the Rule of Five actually asked, the dataset behind it, and when to switch each filter off.

21

PAINS, REOS and structural alerts

Where the substructures came from, the assay mechanisms behind frequent hitters, and what to do with a flagged compound other than delete it.

22

Ligand efficiency metrics

LE, LLE, LELP, BEI and SEI: how to choose between a potent, large, lipophilic hit and a weaker, smaller, cleaner one.

23

Bioisosterism and activity cliffsOptional

The replacements you will meet constantly, and the systematic failure of the similar property principle.