liveBook · Manning

Data curation

As explained in Build AI Drug Discovery Pipelines

Data curation is the systematic review, correction, normalization, and organization of data to improve its quality and consistency.

Data curation systematically reviews, corrects, normalizes, and organizes data so analyses use consistent, trustworthy measurements. In drug-discovery datasets, this can matter more than selecting a model. section 3.2.1

A practical curation check for an experimental property includes:

  • Plot the pIC50 distribution and compare its range with scientifically expected values. section 3.2.1
  • Investigate unusual modes or clusters, which may indicate annotation errors such as mixing nanomolar and micromolar units. section 3.2.1

Simulated unit error

Adding 3.0 to every pIC50 value creates a second distribution separated by three units, illustrating how an annotation mistake can become visible during exploratory review.

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Demonstrates schema inspection and distribution review, then stops at identifying possible unit errors rather than correcting records.
Examples
Uses the hERG blockers dataset and a simulated three-unit pIC50 shift to expose a possible unit inconsistency.
What you can do after
Diagnose potential annotation and unit problems with Python-based schema inspection and distribution plots.
How it connects
Connects data quality with annotation errors, pIC50 distributions, and machine-learning modeling.

Also explained in

Compare how 5 books explain data curation →

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.