A modeling framework for uncertainty quantification in filter-plate-based high-throughput screening of the chromatography partition coefficient

High‑throughput screening (HTS) of chromatography partition coefficients (Kp ) is widely used in protein purification process development, spanning applications with markedly different precision requirements. Despite this widespread use, quantitative guidance on how experimental uncertainty propagates through filter-plate-based K$_p$ assays and informs key design decisions has been lacking. We present an experimentally calibrated Monte Carlo (MC) framework for uncertainty propagation in filterplate-based Kp HTS, which quantifies overall uncertainty and identifies the dominant experimental error sources across purification‑relevant equilibrium regimes. The model is parameterized by characterization of the major sources of experimental errors in a monoclonal antibody (mAb) monomer case study on POROS XS cation‑exchange media and accurately reproduces the magnitude, heteroscedasticity, and plate‑to‑plate variability observed.
Variance decomposition identifies uncertainty in resin slurry distribution and supernatant concentration measurements as the dominant contributors, accounting for ≈ 97 % of total Kp variance over the investigated
... mehr
range of ideal partition coefficients (Kp,ideal). The framework further enables quantitative evaluation of design
choices: analysis of the liquid‑to‑solid phase ratio (β) reveals a clear trade‑off between measurement precision
and protein material demand, allowing derivation of application‑specific phase ratio recommendations.
Overall, this work provides a quantitative basis for uncertainty‑aware HTS design and more reliable Kp data
for purification process development decisions and modeling.