liveBook · Manning

Drug-likeness

As explained in Build AI Drug Discovery Pipelines

Drug-likeness is the degree to which a compound has physicochemical and structural characteristics associated with suitability as a drug.

Drug-likeness estimates whether a compound has physicochemical characteristics associated with oral-drug suitability. It is commonly used as an inexpensive early filter before ligand-based screening, docking, or molecular-dynamics simulations reduce a library further. section 2.1.1 section 2.3.1

A common starting point is Lipinski’s Rule of Five, which flags increased absorption risk when a molecule exceeds more than one threshold:

These rules approximate oral bioavailability rather than proving drug quality. A violation does not automatically imply poor solubility or permeability; project-specific risk thresholds determine how strictly to apply the filter. section 2.1.1 section 2.3.1

Applying an Ro5 filter

For each compound, count violations of the four thresholds, mark it compliant when the count is at most one, and separate compliant from violated records. A library example reduces 212,670 compounds to 199,640 post-filter records. section 2.3.1

Figure 2.1 Virtual screening pipelines are often hierarchical, proceeding with a broad, large number of compounds through simpler, less restrictive, and computationally cheaper processes to whittle down the quantity of compounds. If the 3D structure of the target is not defined, we’re restricted to ligand-based virtual screening. If it is defined, then we can explore structure-based virtual screening, potentially as a downstream component to ligand-based virtual screening.

Use the hierarchical pipeline view to place drug-likeness filtering before more restrictive and computationally expensive screening stages. section 2.1.1

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Defines four Rule-of-Five thresholds and implements counting, but stops short of comprehensive ADMET or pharmacokinetic modeling.
Examples
Demonstrates Rule-of-Five filtering with compound-library statistics and comparisons of molecular-weight, logP, and hydrogen-bond distributions.
What it assumes
Assumes basic compound-property data while defining lipophilicity, polarity, hydrogen-bond donors, and acceptors.
What you can do after
Operate/configure a Rule-of-Five filter that counts property violations and partitions a compound library for downstream screening.
How it connects
Connects oral bioavailability, molecular weight, lipophilicity, polarity, hydrogen bonding, ADMET, virtual screening, docking, and molecular dynamics.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.