Revision of USP Chapter <1039> Chemometrics Published for Comments

In Pharmacopeial Forum 51(4), the USP has released a major revision of General Chapter <1039> Chemometrics for public comment.

Chemometrics is defined as the application of mathematical and statistical methods to extract relevant chemical information from complex multivariate data, often generated by analytical techniques such as spectroscopy or chromatography.

The update reflects the rapid advancement of chemometrics, particularly through developments in machine learning and AI. It aims to support the scientifically sound application of chemometric methods throughout a model’s entire lifecycle and complements existing chapters such as <1010>.

The proposal is based on the version of the chapter that became official on 01 May 2020. According to the briefing notes, key changes include:

• "The 1. Introduction has been edited to reflect the objectives and applications of the discipline.
• Section 2. What is Chemometrics?, has been fully reviewed, reflecting the general overview of the different statistical techniques for modeling, based on three sections, namely 1) unsupervised learning (2.3 Unsupervised Methods), 2) supervised learning with categorical data, and 3) supervised learning with continuous data (2.4 Supervised Methods). Also, a new section describing the types of data structures and the difference of preprocessing methods has been added (2.1.3 Data Structures and 2.2 Signal Processing).
• Section 3. Model Life Cycle has been adapted to be in line with the concept of model life cycle.
• Section 4. Applications of Chemometrics has been updated, reflecting the applications of chemometrics in the pharmaceutical industry."

The draft is available on the Pharmacopeial Forum website. Comments are invited until 30 September 2025.