We are pleased to introduce a special issue of Pharmaceutical Statistics aimed at informing nonclinical biopharmaceutical statisticians and scientists of best statistical practices in the areas of Drug Discovery and Chemicals, Manufacturing, and Controls (CMC). This tutorials issue was organized by the Nonclinical Biostatistics Working Group (NCBWG) of the Biopharmaceutical Section of the American Statistical Association (ASA-BIOP). The NCBWG comprises statisticians from across the pharmaceutical industry whose purposes are to promote the nonclinical biostatistics field through scientific meetings, publication, and educational outreach; broaden the application, reach, and communication of biostatistical methodology and techniques; and foster cooperative efforts among educational, research, and governmental organizations. Originally created in 2003, the NCBWG has sponsored a biennial nonclinical biostatistics conference since 2009 and later joined ASA-BIOP in 2017. For more information on NCBWG, see https://community.amstat.org/biop/workinggroups/ncbwg/index. The tutorials were crafted by expert statisticians from across the industry, who authored 16 works on strategies and approaches with computer code to guide practitioners working in similar fields. We express our gratitude to the many reviewers who dedicated their efforts to refining and enhancing these works, and to the journal of Pharmaceutical Statistics for creating a special issue to feature them.

In discovery, the search for novel drugs and associated mechanisms of action often begins with target identification and target validation, followed by hit-screening to identify compounds that are active against the intended drug target, and culminating with in-vivo animal experimentation to determine the optimal compound for human trials. Eight works in this tutorial issue provide guidance for statisticians and scientists in the in-vitro and in-vivo space. With 13 co-authors from eight departments across four companies, “The partnership between statisticians and the Institutional Animal Care and Use Committee (IACUC)” offers work-related advice to the statistician who works with animal lab scientists. Because in-vivo experiments should be performed with good statistical practices, animals are often randomized into groups, as covered by the work “Randomization in pre-clinical studies: when evolution theory meets statistics”. Sample size and data-analysis considerations for in-vivo dose-response and for mouse tumor-growth studies are respectively provided in “Strategy for designing in-vivo dose-response comparison studies” and “Experimental design considerations and statistical analyses in preclinical tumor growth inhibition studies”. Animal biomarker studies are often used to translate medical knowledge from an animal model to a human. For this type of analysis, we offer “Tutorial on Firth’s logistic regression models for biomarkers in preclinical space”. For assay development for hit-screening campaigns, “Quality by design for preclinical in vitro assay development” offers CMC assay-development philosophy in a discovery space. After the assay is developed, one might find guidance in “Estimating the strength of binding affinity via delta-delta-G for hit screening after a Deming regression calibration”, a collaborative industry/academia work. Finally, for a thorough tutorial on drug-combination analysis, this issue includes “Synergy detection: a practical guide to statistical assessment of potential drug combinations.”

In the CMC space, studies are conducted to optimize the process of manufacturing of drug substance and drug product, to ensure that the process is reproducible and producing a drug product that meets with specifications. To get an understanding of CMC statistics, we open this section with “The role of CMC statisticians: Co-practitioners of the scientific method”. On the path towards developing the drug product manufacturing process, one set of co-authors considers mixture experiments in “Mixture experimentation in pharmaceutical formulations: A tutorial”. Drug substance, drug product, and impurities must be measured with assays that have been thoroughly characterized and understood. For biologics, guidance on immunogenicity testing is provided in “Statistical tutorial for cut-point determination in immunogenicity studies”. Such testing is performed with a qualified and validated immunoassay, where immunoassay performance is provided in “Introduction to qualification and validation of an immunoassay”. Other bioassays also undergo a validation process with guidance provided in “Potency assay variability estimation in practice”. The mean of a process and variance components, such as those due to the assay and the process, are measured against specification limits. A look at a quality statistic that links the mean and total variance to specification limits is given in “Predictive Ppk calculations for biologics and vaccines using a Bayesian approach – a tutorial”. Another approach to process quality is via modeling and simulation against specifications given in regulatory guidance, as given in “Statistical approaches to evaluate in-vitro dissolution data against proposed dissolution specifications”. Finally, for a treatise of machine learning applications to CMC problems, the tutorial edition is rounded out with “What they forgot to tell you about machine learning with an application to pharmaceutical manufacturing”.

For statistician and non-statistician alike, this tutorial special issue aims to bring you both delight and new insights. These works represent the knowledge and techniques of the NCBWG. We hope that these tutorials will spur more interaction with nonclinical statistics teams in industry, academia, and regulatory. We conclude by wishing the reader a successful journey through the tutorials and a second hope that these manuscripts will excite, motivate, and stimulate the readership to new heights in research and development.