Útmutatóhoz kapcsolódó kvíz · Kemény Szint

ADMET and Toxicity Prediction Quiz

Interpret ADMET endpoints, QSAR validation, assay variability and applicability domains without treating predictions as proof of safety.

Kapcsolódó útmutató utakAdmet And Toxicity Prediction
kérdés 1 -ból 8

What does ADMET stand for?

Tekintse meg az összes kérdést8 kérdéseket

A válaszok és magyarázatok megjelennek a kvíz kitöltésekor.

  1. What does ADMET stand for?
    • Absorption, distribution, metabolism, efficacy, and transport
    • Activity, dose, mechanism, efficacy, and testing
    • Absorption, distribution, metabolism, excretion, and toxicity
    • A single clinical safety score
  2. Why are ADMET predictions endpoint-specific?
    • Every assay measures the same safety result
    • Different assays measure distinct biological properties
    • Molecular fingerprints determine all outcomes directly
    • Toxicity can be inferred from solubility alone
  3. Why can a random molecular split overstate performance on new chemical scaffolds?
    • It prevents training examples from being used
    • It guarantees external validation
    • It removes assay noise
    • Related analogs may appear in both training and test sets
  4. What does an applicability domain describe?
    • The chemical and data conditions where the model has relevant training support
    • The set of molecules with the highest predicted safety scores
    • The dataset size used to train the model
    • A preselected safety cutoff for changing a probability into a class
  5. Why document assay source and protocol with training labels?
    • It guarantees every value is noise-free
    • It removes the need for external testing
    • Experimental methods and units can change what a measured value means
    • It determines molecule novelty
  6. What does a positive predicted safety result establish?
    • Proof that a compound is safe in humans
    • A model-based hypothesis that may guide testing
    • A replacement for toxicology assays
    • A guarantee across every dose
  7. Which split can better assess generalization to novel scaffolds?
    • A random split with close analogs in both sets
    • A training-only score
    • A split made after viewing test outcomes
    • A scaffold-separated split
  8. Which information supports transparent QSAR use?
    • Defined endpoint, model algorithm, applicability domain and validation evidence
    • Only a high training score
    • Only the molecule string
    • A single combined ADMET label