Statistics in Biopharmaceutical Research

Advanced Computational and Statistical Learning Methods for Pharmaceutical Research: Focusing on Reproducibility, Robustness, Integrity, Interpretability and Other Desired Properties

Submission Timeline
First submission: June 1, 2026
Submission deadline: November 1, 2026

Statistics in Biopharmaceutical Research (SBR) invites submissions for a special issue on computational methods with desired properties for biopharmaceutical applications. The biopharmaceutical industry faces increasingly complex challenges, such as leveraging real-world data (RWD) from electronic health records and registries, analyzing high-dimensional clinical trial data, modeling drug efficacy and safety, and optimizing manufacturing processes. With the growing volume and diversity of biomedical data, advanced computational methods—including artificial intelligence (AI) and machine learning (ML)—have become strong candidates for extracting actionable insights, enhancing predictive accuracy, and supporting data-driven decision making. If properly implemented, these modern techniques may accelerate drug development timelines, may enable more precise patient stratification, and may inform adaptive trial designs.

SBR has no submission or publication charges.

Focusing on biopharmaceutical applications, those computational methods should satisfy several desired properties, such as reproducibility (the ability to replicate results under the same setting), robustness (the ability to draw a consistent conclusion under a varying range of assumptions), integrity (the ability to deliver trustworthy and traceable results) and interpretability (the ability to effectively communicate findings with a broader audience). This special issue seeks to highlight the importance of those properties when applying traditional, modified or innovative computational methods to biopharmaceutical research.

Scope:
Several types of papers are welcomed in this special issue:

1. Propose an innovative computational approach with several desired properties (e.g., reproducibility, robustness, integrity, interpretability) for biopharmaceutical applications, in addition to superior performance (e.g., power gain).
2. Modify an existing computational approach to satisfy more desired properties for biopharmaceutical applications.
3. Provide insightful points to consider and constructive mitigation strategies for implementing existing computational approaches in biopharmaceutical applications with respect to those desired properties.

The list of desired properties is general to include other aspects, such as computational efficiency, simplicity, data privacy and ethicality. Taking the computational efficiency as an example, authors can propose a shortcut to an existing method with heavy computations, or propose a more efficient simulation design, or include advanced computational techniques to reduce computational time, or provide insights on patient-level or summary-level simulations with privacy-preserving considerations.
Potential topics include, but are not limited to,

• Bayesian computation
• Causal inference with artificial intelligence and machine learning
• Deep neural networks
• Digital / synthetic twins
• Integration of pharmacometrics and statistical modeling
• Large language models
• Numerical optimization
• Resampling methods
• Targeted learning

Papers should be motivated by important biopharmaceutical applications, such as RWE (real-world evidence), integration of RWE, clinical and omics data, complex and innovative clinical trial designs, precision medicine, pharmacovigilance, early-phase clinical trials, HTA (health technology assessment), decentralized clinical trials, digital endpoints, CMC (chemistry, manufacturing and controls), and operational aspects of clinical trials (e.g., recruitment monitoring, site monitoring, and clinical drug supply).
Accepted papers are strongly recommended to include computational code, associated data and proper documents to reproduce results. Rationales for not doing so should be explained to the editorial team.