liveBook · Manning

Lipinski's rule of five

As explained in Build AI Drug Discovery Pipelines

The Lipinski rule of five is a heuristic set of molecular-property guidelines for estimating a compound's oral drug-likeness.

Lipinski’s Rule of Five is a quick filter for estimating whether a small molecule may be suitable for oral drug development. It uses molecular properties related to aqueous solubility, intestinal permeability, and absorption. A compound is flagged when it violates more than one threshold; a violation indicates increased risk, not guaranteed failure. section 2.1.1 section 2.3.1

Check these four molecular-property limits:

  • Keep molecular weight below 500 daltons. section 2.1.1
  • Keep calculated lipophilicity, or logP, below 5; higher values can reduce dissolution in aqueous digestive fluids. section 2.3.1
  • Keep hydrogen bond acceptors at 10 or fewer and hydrogen bond donors at 5 or fewer. section 2.3.1
  • Treat the filter as a starting point for small, orally administered molecules rather than a universal definition of drug-likeness. section 2.1.1

Filtering a compound library

A dataframe filter can count violations in ExactMolWt, MolLogP, NumHDonors, and NumHAcceptors, then retain compounds with at most one violation. Applied to the SPECS library, the illustrated workflow reduced 212,670 compounds to 199,640 post-filter compounds. section 2.3.1

Listing 2.4 Applying an Ro5 Filter
def apply_lipinski_filter(df):
    def count_ro5_violations(row):
        violations = 0
        if row['ExactMolWt'] > 500: violations += 1
        if row['MolLogP'] > 5: violations += 1
        if row['NumHDonors'] > 5: violations += 1
        if row['NumHAcceptors'] > 10: violations += 1
        return violations
    
    df['ro5_violations'] = df.apply(count_ro5_violations, axis=1)
    df['ro5_compliant'] = (    #A
        df['ro5_violations'] <= 1
    )
    
    compliant_df = df[df['ro5_compliant']]
    violated_df = df[~df['ro5_compliant']]
    return df, compliant_df, violated_df
 
specs, specs_ro5_compliant, specs_ro5_violated = \
    apply_lipinski_filter(specs)

Inspect the runnable Ro5 filtering pattern, including violation counting, compliance labeling, and separation of retained and rejected compounds. section 2.3.1

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Defines four thresholds, explains their absorption rationale, and demonstrates dataframe implementation while stopping short of universal drug-likeness claims.
Examples
Uses a SPECS compound-library filter, summary statistics, and molecular-weight, hydrogen-bonding, and logP distributions.
What it assumes
Defines molecular weight, logP, polarity, hydrogen bond donors, and hydrogen bond acceptors before applying the filter.
What you can do after
Apply it via a dataframe filter to count violations and separate compounds for further screening.
How it connects
Connects oral bioavailability with absorption, permeability, solubility, lipophilicity, hydrogen bonding, ADMET, and virtual screening.

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.