PQn

PQN is an acronym used in multiple disciplines, referring to different concepts depending on the context. In metabolomics, PQN stands for Probabilistic Quotient Normalization, a data normalization method designed to reduce the impact of sample dilution on spectral intensities. In practice, a reference spectrum is chosen (often the median spectrum across samples). For each sample, the quotients of its peak intensities to the corresponding reference values are computed; the median of those quotients estimates the dilution factor for that sample, and all intensities are scaled by this factor. PQN helps improve comparability across samples and is robust to outliers, making it widely used in mass spectrometry and NMR-based metabolomics workflows.

In numerical optimization, PQN commonly refers to Projected Quasi-Newton methods, including variants that handle bound constraints. These methods maintain a positive definite approximation to the inverse Hessian and perform line searches along projected directions to keep iterates within the feasible region. For large-scale problems, limited-memory variants of PQN are employed to reduce storage and computational demands while still offering efficient convergence.

The exact meaning of PQN is context-dependent, and the term may have other interpretations in specialized fields. Readers encountering PQN should consult domain-specific literature or documentation to confirm the intended concept.