ADMET prediction and in silico property analysis MCQs With Answer

ADMET prediction and in silico property analysis MCQs With Answer

by

Introduction: This quiz set on ADMET prediction and in silico property analysis is designed for M.Pharm students preparing for MPC 203T Computer Aided Drug Design. It focuses on practical and theoretical aspects of predicting Absorption, Distribution, Metabolism, Excretion and Toxicity using computational tools. Questions emphasize molecular descriptors, rule-based filters, machine learning approaches, common software platforms, in vitro–in silico correlations and interpretation of ADMET outputs for lead optimization. The aim is to strengthen exam preparation and decision-making skills in early drug discovery by testing knowledge of models, limitations, and how to integrate in silico ADMET into rational compound design.

Q1. Which property is most directly assessed by the octanol–water partition coefficient (logP) in drug design?

  • Lipophilicity affecting membrane permeability
  • Aqueous solubility
  • Metabolic stability toward CYP enzymes
  • Cardiac hERG blockade potential

Correct Answer: Lipophilicity affecting membrane permeability

Q2. Which rule of thumb defines limits for molecular weight, hydrogen bond donors, hydrogen bond acceptors and logP for likely oral drugs?

  • Veber rule
  • Ghose filter
  • Lipinski’s Rule of Five
  • Fsp3 guideline

Correct Answer: Lipinski’s Rule of Five

Q3. Topological polar surface area (tPSA) below which approximate value is commonly associated with better blood–brain barrier penetration?

  • Less than ~90 Ų
  • Less than ~150 Ų
  • Greater than ~140 Ų
  • Between 200–300 Ų

Correct Answer: Less than ~90 Ų

Q4. Which specialized software is primarily used to predict likely sites of metabolism and CYP-mediated metabolism on a molecule?

  • MetaSite
  • SwissADME
  • ROCS
  • AutoDock Vina

Correct Answer: MetaSite

Q5. Which descriptor specifically counts hydrogen bond donors in a molecule?

  • HBD (Hydrogen bond donors)
  • HBA (Hydrogen bond acceptors)
  • tPSA (Topological polar surface area)
  • LogS (Aqueous solubility)

Correct Answer: HBD (Hydrogen bond donors)

Q6. The Caco-2 cell assay is an in vitro model most commonly used to estimate which ADME property?

  • Intestinal permeability
  • Renal clearance
  • Plasma protein binding
  • Hepatic intrinsic clearance

Correct Answer: Intestinal permeability

Q7. Which machine learning algorithm is well-suited to capture nonlinear relationships in QSAR/ADMET modelling?

  • Random forest
  • Ordinary least squares linear regression
  • Principal component analysis (PCA)
  • k-means clustering

Correct Answer: Random forest

Q8. In silico prediction of which endpoint is considered a primary indicator of potential cardiotoxicity?

  • hERG channel inhibition
  • AMES mutagenicity
  • Hepatotoxicity prediction
  • Plasma protein binding percentage

Correct Answer: hERG channel inhibition

Q9. Physiologically based pharmacokinetic (PBPK) models are mainly used to predict which aspect of a drug?

  • Whole-body time-dependent concentration profiles
  • Crystal packing and solid form
  • Protein–ligand docking scores
  • pKa from first principles

Correct Answer: Whole-body time-dependent concentration profiles

Q10. Which limitation is most commonly cited for in silico ADMET models when applied to novel chemotypes?

  • Applicability domain constraints (out-of-domain predictions unreliable)
  • Excessive experimental throughput
  • Unlimited interpretability of complex models
  • Guaranteed human clinical accuracy

Correct Answer: Applicability domain constraints (out-of-domain predictions unreliable)

Q11. Which combination of properties most directly determines oral bioavailability in early drug design?

  • Aqueous solubility and membrane permeability
  • hERG inhibition and AMES mutagenicity
  • Protein tertiary structure and melting point
  • Caco-2 assay and docking score to target

Correct Answer: Aqueous solubility and membrane permeability

Q12. Which online tool is widely used for in silico acute toxicity class prediction (including LD50 categorization)?

  • ProTox-II
  • AutoDock
  • ChemDraw
  • MM-GBSA

Correct Answer: ProTox-II

Q13. Which molecular descriptor best quantifies molecular flexibility relevant to oral absorption?

  • Number of rotatable bonds (rotB)
  • Topological polar surface area (tPSA)
  • Fraction of aromatic atoms
  • pKa of the most acidic group

Correct Answer: Number of rotatable bonds (rotB)

Q14. P-glycoprotein (P-gp) expression in epithelial barriers primarily affects a drug by:

  • Acting as an efflux transporter that reduces intracellular concentrations and can limit brain penetration
  • Directly inhibiting CYP3A4 metabolic activity
  • Increasing plasma protein binding dramatically
  • Altering compound crystallinity

Correct Answer: Acting as an efflux transporter that reduces intracellular concentrations and can limit brain penetration

Q15. Which in silico approach is commonly used to estimate aqueous solubility from molecular structure?

  • QSPR (Quantitative Structure–Property Relationship) models
  • X-ray crystallography
  • Molecular cloning
  • Surface plasmon resonance

Correct Answer: QSPR (Quantitative Structure–Property Relationship) models

Q16. Which cytochrome P450 isoform is responsible for metabolism of the largest fraction of marketed drugs?

  • CYP3A4
  • CYP2D6
  • CYP1A2
  • CYP2E1

Correct Answer: CYP3A4

Q17. Veber’s rule emphasizes which pair of descriptors to predict good oral bioavailability?

  • Topological polar surface area and number of rotatable bonds
  • Molecular weight and logP
  • HBD and HBA counts only
  • Aromatic ring count and fraction sp3

Correct Answer: Topological polar surface area and number of rotatable bonds

Q18. What is a principal advantage of ensemble learning methods (e.g., random forest, gradient boosting) in ADMET prediction?

  • They reduce overfitting and typically improve predictive performance by combining many models
  • They always produce single-descriptor mechanistic explanations
  • They are guaranteed to extrapolate well outside training data
  • They eliminate the need for data curation

Correct Answer: They reduce overfitting and typically improve predictive performance by combining many models

Q19. Which tool or resource is specifically developed for in silico prediction of hERG blockade?

  • Pred-hERG
  • PROPKA
  • BLAST
  • MAFFT

Correct Answer: Pred-hERG

Q20. How is the “applicability domain” of an ADMET model best described?

  • The chemical space and descriptor range within which model predictions are considered reliable
  • The list of software programs used to build the model
  • The CPU time required to run the model
  • The target protein binding pocket volume

Correct Answer: The chemical space and descriptor range within which model predictions are considered reliable

Leave a Comment