The Smart Benefit of Digital Systems in Preclinical Studies

Published on: 
BioPharm International, BioPharm International, May 2023, Volume 36, Issue 05
Pages: 30–33

Preclinical studies specifically benefit from the “smart” drug development trend through deeper data access and analysis.

The advancement of “smart” drug development reaches far back in the drug development cycle, starting at the R&D phase and moving through the pipeline through to clinical studies. Data gathering and deep data analysis in the preclinical stages of development are growing increasingly crucial as the biopharmaceutical industry strategizes ways to better decrease costs while speeding up timelines to market. Here, a closer look at smart tools highlights how they support enhanced and intelligent data gathering as well as analysis in drug discovery and preclinical studies.

One way to gain speed in drug development is through the appropriate management of preclinical trials. The better that preclinical data are managed within organizations and across organizations, the higher the probability of getting therapeutics to patients faster. For instance, having digital systems that are designed in such a way that they simplify the collection, organization, visualization, and sharing of preclinical data can significantly advance progress in drug discovery and development (1).

Another example of the use of smart tools in early stages of drug development involves artificial-intelligence (AI)-based predictions. AI-based predictions can guide researchers as to which animal models (e.g., which rat experiments) should be conducted for meaningful preclinical data collection. Such predictive modeling can also explain discrepancies and variabilities observed in preclinical studies, which can elucidate important disease pathways (1).

To further explore the intricacies of smart tool utilization in preclinical studies and early phase drug development in general, BioPharm International® spoke with Jijie Gu, PhD, chief scientific officer and executive vice-president of WuXi Biologics.

Early phase tools

BioPharm: Tools such as AI, automation, and machine learning are being utilized for bio/pharmaceutical manufacturing processes, but how are they being utilized in the R&D/drug discovery/drug screening phases?

Gu (WuXi Biologics): Automation and digitalization are revolutionizing the bio/pharmaceutical industry and are being utilized across all phases, including drug discovery.

Similar to manufacturing processes, drug discovery also utilizes these technologies to enhance efficiency, minimize cost, and accelerate drug discovery. Automation forms the foundation of these advances. [M]any discovery processes have been automated and streamlined [for instance, within the WuXi Biologics’ discovery service department], shortening the research timeline by months. For instance, our single B cell sorting platform (Beacon) automatically screens a large amount of plasma B cells from immunized animals or recovered patients and identifies promising antibody lead candidates in days. These antibodies can then be rapidly produced on [a] fully automated protein production platform and characterized on [a] high-throughput screening platform. The functional leads, if derived from non-human sources, are then humanized through automatic design—in our case, via our online algorithm—in seconds, before undergoing further in-vitro/in-vivo characterizations.

In addition to efficiency improvement, digitalization and machine learning tools are also empowering innovation in our discovery services process. We leverage IoT [Internet of Things] devices to gather and analyze real-time data generated by automations. Our use of digitized tools extends to analyzing proteomics and pharmacology data to identify novel targets and more effective therapeutic strategies for our clients. Additionally, we have leveraged machine learning to predict and optimize candidate antibody properties, including stability, solubility, chemical properties, and binding affinity, thereby enhancing their therapeutic efficacy. We are also developing de novo antibody design algorithms to augment our capability for discovery services.

BioPharm: Are digitalization tools such as these important for preclinical testing of drug candidates, and how so?

Gu (WuXi Biologics): Yes, these tools are very important for testing drug candidates in the preclinical phase. Extensive testing of drug candidates is necessary to determine their potential for development, efficacy, and safety, before they can progress to human clinical trials.

At WuXi Biologics, automation significantly expedites preclinical testing by enabling high-throughput screening of a large number of drug candidates. This approach provides a more efficient and comprehensive way to identify the most promising drug candidate. Automation also reduces the occurrence of human error and improves the reproducibility of tests, resulting in a more precise evaluation of drug candidates. Our digitalization system helps store, manage, and analyze vast amounts of data generated during preclinical testing, including drug biophysical properties, in-vitro/in-vivo function, pharmacokinetics, toxicity, and more. We also develop digital tools for efficient data analyzation and visualization, facilitating the identification of trends or patterns in the preclinical drug profiling.

Additionally, we utilize machine learning informed by historical preclinical/clinical testing data to predict new drug candidates with a higher probability of success in clinical trials, or to identify potential risks in their developability, efficacy, or safety. If concerns arise, digitalized tools are adopted to optimize drug properties such as affinity and stability to enhance its performance in preclinical testing.

Preclinical benefit


BioPharm: Why is it important for the ‘smart’ drug concept that tools such as these be brought in during the early stages of drug development, such as in preclinical studies?

Gu (WuXi Biologics): Such digitalization tools help boost the bandwidth and depth of preclinical drug development. For instance, at WuXi Biologics, by employing digitized and automated molecule generation, it is possible to generate more diverse candidate molecules at a higher quantity, so that likelihood of success in selecting a molecule with clinical potential is significantly improved. In addition, by applying novel digitalized analysis to mechanistic studies, it is possible to understand differences between competing molecules with greater granularity. As a result, truly differentiated molecules with distinct mechanisms of action can be discovered, improving patient quality of life.

BioPharm: Are you seeing an enhancement in preclinical data generation and data analysis as a result of using digitalization tools? Or, if digitalization tools are not necessary in preclinical studies, why is that so?

Gu (WuXi Biologics): Yes. Digitalization tools have greatly enhanced both the capacity and efficiency of data generation and data analysis.

At WuXi Biologics, digitalization coupled with automation has greatly increased the throughput of traditional assays such as ELISA [enzyme-linked immunosorbent assay], FACS [fluorescence-activated cell sorting], SPR [surface plasmon resonance], etc. A significant number of candidate molecules can therefore be screened at the same time, increasing the likelihood of discovering the preclinical candidate molecule for the greatest potential of clinical success.

Digitalization has also introduced new methods of analyzing preclinical experimentations. For instance, digitalization and the accompanying computerized image analysis has greatly improved resolution and accuracy of pathology analysis. Next-generation sequencing, together with innovative algorithms, has also given us the ability to look at in-vitro and in-vivo pharmacology data in dimensions unimaginable in the past. By employing such advanced analyses, WuXi Biologics’ discovery organizations have developed an ability to discover differentiated molecules with distinct mechanisms of action.

Clearing up misconceptions

BioPharm: What misunderstandings exist in the bio/pharmaceutical industry pertaining to preclinical studies, and what tips would you give to clarify any misunderstanding?

Gu (WuXi Biologics): One big misunderstanding is about the role and implication of preclinical efficacy and safety studies in animal models.

Human diseases usually have vastly complex and diversified etiologies with enormous variations across the patient population. However, each animal model generally has a specific etiology, a certain set of dominant pathogenic factors and a defined disease course. In a sense, each animal model may only resemble certain mechanism(s) of a complex disease represented in [a] certain patient population of a given disease. Therefore, efficacy in any certain model is not a guarantee of clinical success in the wider human patient population.

However, preclinical studies do indeed have great value in evaluating the capability of therapeutic agents in modulating any certain critical pathway. Therefore, for diseases where a significant population of patients share a dominant pathologic pathway, such as psoriasis with the IL-23-Th17-IL-17A [interleukin-23-T-helper 17-interleukin-17A] axis, a model where these factors are dominant can have considerable predictive power.

For diseases with more complex pathologies, it is importance to have a deep understanding of disease biology and have multiple complementary models to cross-validate the therapeutic potential of any agent. Novel tools and aforementioned methods can also help develop a deeper understanding of mechanisms of action and differentiation within existing treatments.


1. Mirasol, F. Using Smart Tools for Smart Development. Pharm. Technol. 2023, 47 (5), 16–20.

About the author

Feliza Mirasol is the science editor for BioPharm International.

Article Details

BioPharm International
Volume 36, No.5
May 2023
Pages 30–33


When referring to this article, please cite it as Mirasol, F. The Smart Benefit of Digital Systems in Preclinical Studies. BioPharm International 2023, 36 (5), 30–33.