Accelerating drug discovery with AI innovation

With SAS Viya, Shionogi developed a system to semiautomate clinical trials without compromising safety, efficacy or compliance.

In drug development, every minute lost risks lives and diminishes hope. That thought haunted the data scientists at Shionogi, a leading Japanese pharmaceutical company renowned for its contributions to researching infectious diseases, the nervous system and cancer pain management. Surely, AI, data science and the innovative problem solvers in Shionogi's Data Science Department could compress the yearslong clinical trials process required for drug development – and do it without compromising safety or efficacy.

They were right. The team invented a novel AI-based programming system using SAS Viya. The system is redefining drug development timelines and reshaping the pharmaceutical industry.

Cutting through the noise

Like most pharmaceutical companies, Shionogi has long trusted the SAS platform for regulatory-compliant clinical trial analysis and reporting.

A common bottleneck is the pharmaceutical industry’s practice of creating new analytical programs for every clinical trial – even when processes and tasks are repeated from trial to trial. These routine tasks require hundreds of hours for each clinical trial and consume tens of thousands of hours a year across Shionogi’s drug portfolio.

According to Dr. Yoshitake Kitanishi, Associate Corporate Officer and Head of the Data Science Department at Shionogi, the team needed AI technology to automate routine tasks so they could reclaim the lost time for more groundbreaking advancements. “With AI automation, we can delegate tasks that should be done by machines to the machines, freeing up time for humans to focus on tasks humans should do,” Kitanishi explained.

Enter the AI SAS Programmer System (AI-SAS), an innovation powered by SAS Viya.

The breakthrough: Building AI-SAS on SAS Viya

When conceptualizing the AI-SAS system, the team was looking to solve three challenges: simplify the implementation of machine learning and deep learning capabilities, ensure robust data governance and transform the business from a focus purely on drug discovery to a global health care-as-a-service (HaaS) provider. And they wanted to accomplish that from a single platform.

As the name implies, AI-SAS is a trailblazing system that semiautomatically generates SAS programs using AI. It uses various machine learning and deep learning techniques, such as convolutional neural networks (CNNs), which are generally implemented in Python. Python programming is complex, and there were concerns that Python would make the system a black box for Shionogi programmers, the majority of whom are SAS users.

“Python in the SAS Viya interface produced highly readable programming with a very simple description,” explained Kitanishi. “Other machine learning tasks could be implemented on the SAS Viya interface as well, which helped us simplify the system’s construction.”

The result? The standard analysis time per trial was reduced by 30% from 350 hours. "AI-SAS has significantly reduced the amount of work required to analyze clinical trials and improved their efficiency," Kitanishi added.

A catalyst for the industry’s transformation

The impact of AI-SAS goes beyond operational efficiency. The system is the foundation for Shionogi's broader mission to convert its go-to-market business model to a HaaS provider.

AI-SAS’ semiautomation will help address global medical challenges more quickly than ever before. Not only that, but the time saved by faster analysis times will spur more challenging and creative endeavors in the health care industry.

"AI-SAS is now being adopted by other Japanese pharmaceutical companies and CROs,” explained Kitanishi. “We intend to spread in Japan and globally.”

Results that resonate

The numbers speak for themselves. AI-SAS reduced analysis time by 30%, helping establish Shionogi as a leader in digital transformation. Certified as a "Noteworthy DX Company" by the Ministry of Economy, Trade and Industry (METI), Shionogi is proof that AI can bring lasting benefits to society.

Shionogi's success with AI-SAS is just the beginning. The company plans to expand its AI approach into other areas, such as human resources, health management and real-world data analysis. “Our goal is to empower companies worldwide to deliver new patient therapies with unprecedented speed and precision,” Kitanishi said.

Kitanishi’s advice to other data science teams in the pharmaceutical industry? “SAS is the gold standard analysis tool – trusted for its highly accurate analytical results,” Kitanishi said. “If you want to generate high-quality evidence, SAS is indispensable and should be placed at the center of your work.”