Artificial intelligence and a flood of data in the pharmaceutical industry will likely change some of the current functions of its data scientists, experts say, but the ability to learn and adapt to new technologies will remain key in this role.
Pictured: Woman with Vintage Computers/Taylor Tieden for BioSpace
Data analysis in the biopharma industry is key to decision-making at all levels: during the drug discovery process; in clinical trials, from patient screening to the assessment of safety and efficacy of compounds; and even during the marketing period to position a product competitively.
The importance of data in pharma has increased over time as the approach to medicine more broadly has changed, Jenny Siferd, director of talent attraction at Asklepios BioPharmaceutical, Inc. (AskBio), told BioSpace. “Part of the reason why [data analysis] is becoming so important is probably, in my opinion, just the shift over the years to more of a focus on precision medicine and targeted therapies.” Obtaining more data on patient characteristics and drug responses may help to determine who will respond best to a therapy, she said, while understanding and translating such data could lead to safer drugs.
Recent technological developments have also played a part in this explosion of data. Wearables, for instance, can take people’s heart rates and blood pressure 24 hours a day, seven days a week, generating much more data compared to a more traditional visit to a hospital to have vital signs measured, pointed out Graham Clark, the CEO of contract research organization Phastar. “The volume and the breadth and the different types of data that are available now are massive and growing exponentially,” he said.
This increasing role for data in pharma has translated into a greater demand for staff who are capable of handling and analyzing it. The “role has sort of grown in terms of importance within the pharmaceutical world,” Siferd said. This growth mirrors what is happening across the board in many sectors. The U.S. Bureau of Labor Statistics has predicted, for instance, that jobs for data scientists will grow by 36% between 2021 and 2031.
Data analysis is a hot subject, Siferd added, and university “students are in tune with that, and they understand that this field in general is exploding.” At the bachelor’s level, the number of degrees in data science awarded in 2022 was 10 times that of 2020, according to the National Center for Education Statistics. This strong interest of students in data analysis is “something that has changed from let’s say, 10 years ago or 20 years ago,” she said.
AI and Data Science
In addition to data volume, another driver that could potentially shape the role of the data analysts in pharma is the growing use and sophistication of artificial intelligence (AI). “We see a number of use cases where AI can be used to speed up validation processes, checking for inconsistencies and errors and problems and anomalies in datasets,” Clark said.
He added that it may also be used to mine historical events to project future developments. In an industry where developing a new drug takes years and can cost upwards of $2 billion, these AI capabilities could be hugely valuable.
The use of AI has yet to substantially change the role of data scientists in pharma, Siferd said, but “I think that’s certainly coming.”
Moreover, pharmaceutical companies’ use of AI may not rely primarily on their own data analysts. The industry could also partner with other teams on AI applications, suggested Clark, whose company offers data analysis and data science services.
The pharma industry “doesn’t have embedded within it necessarily all of the skills and capabilities that it’s going to need to be able to utilize AI effectively,” he said. “How players within the industry collaborate with partners and software providers, how they access the right level of expertise and, in particular, the right caliber of people to be able to take them on the journey is going to be really important.”
Same Skills, New Tools
Whatever the precise reasons for the shift, biopharma has moved in recent years toward more stringent educational requirements for applicants for data-related roles, experts told BioSpace. While a Master’s or a Ph.D. was helpful years ago for building a career in data science, a postgraduate degree is now “pretty much required,” Siferd said.
Bryan Wells, executive recruiter and founder of Sanford Rose Associates affiliate Dark Horse Talent who works with drug discovery and biomedical research teams, concurred. Biopharma companies are mostly looking for people with a Ph.D. when filling data-related roles, he said.
Traditional skills such as programming and other computational competencies will continue to be key in these roles. “Right now, Python and R are the two kind of hot coding languages that we would look for, and then experience with modeling, statistical applications, deep data platforms, statistical analysis of large datasets and really pulling that together and creating dashboards and visualizations,” Siferd said.
Clark said that while traditional skills aren’t going to go away, the ability to learn new tools will need to sit alongside “competencies that allow people to utilize other pieces of software, . . . the ability to be able to train and interpret AI models [and] to be able to understand the output from those models.” Wells agreed that companies will likely start “to focus more towards people who have the experience with AI and machine learning.”
“But beyond all of that, I think we really look for people who have just strong analytical skills and problem-solving skills and creative thinking skills,” Siferd added. “The ability to communicate and to really translate the data to what it means to the business is very important as well.”
“You can’t just be in the technology,” Siferd said. “You have to have all of the soft skills around it to really be able to make sense to the business.”
Alejandra Manjarrez is a freelance science writer based in Mexico City. Reach her on her website.
