Machine learning predictions of IgG1 and IgG4 self-association and high-concentration solution properties
Clinical Snapshot
PICO Framework
| P — Population | Panel of clinical-stage IgG1 and IgG4 monoclonal antibodies |
| I — Intervention | Machine learning-based prediction models for self-association behavior |
| C — Comparator | Experimental measurements of self-association, viscosity, and recovery properties |
| O — Outcomes | Self-association behavior, viscosity, antibody recovery, and high-concentration solution properties across different pH and excipient conditions |
Bottom Line
This cross-sectional study developed machine learning models to predict self-association behavior of clinical-stage IgG1 and IgG4 antibodies under various formulation conditions. The research identified acidic pH (5.0) with proline (200 mM) as optimal for reducing self-association, with correlations observed between self-association, viscosity, and recovery properties. While the systematic experimental approach and use of interpretable ML models are strengths, the study lacks quantified effect sizes and detailed selection criteria for the antibody panel. The findings have immediate practical application for pharmaceutical development, potentially accelerating identification of suitable candidates for subcutaneous delivery. However, validation in independent datasets and consideration of additional confounding factors would strengthen the evidence base. The work represents a valuable tool for early-stage antibody optimization but requires further validation before widespread clinical application.
Key Findings
P Value: Not provided
Effect Size: Not quantified in abstract
Primary Outcome: Machine learning models successfully predicted IgG1 and IgG4 self-association behavior
Nnt Or Sensitivity: Acidic formulations (pH 5) with proline (200 mM) most effective at suppressing self-association
Confidence Interval: Not provided
Clinical Application
High - computational models can be integrated into early drug discovery pipelines Relevant to Australian pharmaceutical industry and TGA-regulated therapeutic development; could inform PBS submissions for subcutaneous antibody formulations Pharmaceutical companies developing monoclonal antibody therapeutics for subcutaneous delivery
Abstract
To meet the widespread demand for subcutaneous delivery of antibody therapeutics, candidates with low viscosity, high solubility, and/or low aggregation propensity in concentrated formulations must be identified. Moreover, early identification of candidates with low self-association increases the likelihood of success at later stages of the development process. Here, we experimentally profile the self-association behavior of a panel of clinical-stage antibodies as a function of pH, excipient content, and antibody isotype. We find that acidic formulations (pH 5) with proline (200 mM) are most effective at suppressing self-association for both IgG1 and IgG4 variants. Moreover, our self-association measurements are correlated with antibody viscosity measurements and inversely correlated with antibody recovery after their concentration using membrane filters. Notably, we developed interpretable machine learning-based classifier and regressor models for predicting IgG1 and IgG4 self-association and demonstrated that they identify antibodies with favorable high-concentration properties. These findings are expected to improve the antibody development process by facilitating the identification of drug-like molecules during their discovery and optimization.
References
- 1.Kwon, N. Y., Brown, C. N., Chen, H. T., Cottle, S. R., Kelly, R. M., Jones, B. E., Weiss, W. F., & Tessier, P. M. (2026). Machine learning predictions of IgG1 and IgG4 self-association and high-concentration solution properties. mAbs. https://doi.org/10.1080/19420862.2026.2663641
Related Research
CPT: pharmacometrics & systems pharmacology
Structured Schemas for Provenance-Rich, LLM-Assisted QSP Model Calibration
2 Aug 2026
Journal of pharmacokinetics and pharmacodynamics
Diffusion models for virtual populations and pharmacometric simulations
1 Aug 2026
Journal of chemical information and modeling
HyperDC: A Non-Uniform Hypergraph Framework for Dual- and Higher-Order Drug Combination Recommendation Across Diverse Complex Diseases
29 July 2026
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service