Free PhysChem Calculator

In 1997, Lipinski and colleagues at Pfizer analysed compounds that had reached Phase II clinical trials and identified property ranges correlated with good oral absorption. The Rule of Five flags poor absorption when:

• Molecular weight is greater than 500 Da
• LogP (lipophilicity) is greater than 5
• There are more than 5 hydrogen-bond donors (NH and OH groups)
• There are more than 10 hydrogen-bond acceptors (N and O atoms)

It is called the "Rule of Five" because every threshold is a multiple of five: a useful reminder that these are round-number heuristics, not precise cutoffs. Compounds violating two or more parameters are statistically less likely to succeed as oral drugs, but these are guidelines for probability, not laws of nature.

In 2002, Veber and colleagues at GlaxoSmithKline analysed over 1,100 drug candidates tested for oral bioavailability in rats. They found that molecular flexibility and polarity were strong predictors of oral absorption, independent of molecular weight:

• Topological polar surface area (TPSA) of 140 Ų or less
• Rotatable bonds of 10 or fewer
• Total hydrogen-bond count (HBD + HBA) of 12 or fewer, proposed as an alternative polarity measure to TPSA

TPSA measures the surface area contributed by polar atoms (nitrogen, oxygen, and their attached hydrogens). Higher polarity reduces passive membrane permeability. Rotatable bonds measure molecular flexibility: a more flexible molecule must pay a larger conformational entropy cost to adopt the geometry needed for membrane permeation, which correlates with reduced oral bioavailability. The total H-bond count captures overall polarity in a complementary way to TPSA.

This calculator checks all three Veber criteria alongside Lipinski's Rule of Five and displays all seven properties in a radar chart for easy visual comparison.

QED (Quantitative Estimate of Drug-likeness) was introduced by Bickerton et al. in Nature Chemistry (2012). Instead of applying binary pass/fail thresholds, QED maps each of eight molecular properties through a desirability function derived from the distributions observed in approved oral drugs, then combines them into a single score between 0 and 1.

The eight QED properties are:

• Molecular weight (MW)
• LogP (lipophilicity)
• H-bond donors (HBD)
• H-bond acceptors (HBA)
• Polar surface area (PSA)
• Rotatable bonds
• Aromatic rings
• Structural alerts (116 SMARTS patterns flagging reactive or toxic substructures)

As approximate thresholds, a QED score of 0.65 or above is generally considered favourable, 0.4 to 0.65 is moderate, and below 0.4 is unfavourable. Because QED uses weighted desirability functions rather than hard cutoffs, it captures a more nuanced picture of drug-likeness than any single rule set.

The Ro5 predicts the likelihood of adequate absorption, not overall clinical success. Many approved drugs violate one or more thresholds:

• Atorvastatin (MW 559 Da) exceeds the MW threshold yet is one of the most commercially successful drugs in history.
• Erythromycin (MW 734 Da, ~30% bioavailability), rifampicin (MW 823 Da, ~90% bioavailability), and cyclosporine A (MW 1203 Da, 20-60% bioavailability) have been successful oral medicines for decades.
• More recently, HCV protease inhibitors like glecaprevir (MW 838 Da) and cancer drugs like venetoclax (MW 868 Da) achieve effective oral exposure.

Roughly 27% of oral drugs approved since 2000 exceed 500 Da. Of the four Lipinski parameters, LogP > 5 is arguably the most consequential violation. High lipophilicity correlates with increased attrition across all stages of development.