Digital Biopharma Facing Data Science Skills Gap

Digital Biopharma Facing Data Science Skills Gap

Biopharma’s digital transition has exposed a variety of skills shortages. In response, the industry has built know-how in areas like automation and cybersecurity. But elsewhere, skills are still lacking.

For example, it remains a struggle to find data scientists—those with the ability to draw usable information from raw data—according to Jason Beckwith, PhD, managing director of biopharma recruitment specialist, Evolution Search Partners.

“There seems to be an abundance of mature talent in certain areas like automation, cybersecurity, and sensors, while elsewhere there are specific expertise deficits, notably in data science,” he says.

Beckwith cites integration, analysis, and statistical modeling as well as machine learning as areas of data science where the shortfall is most pronounced.

“The skills gap associated with data science is evident when you look at the relative lack of expertise using technologies such as Apache Spark—a real-time big data processing tool—compared with other industries. Further, NoSQL, a relatively new database management tool, is another area where there is a significant shortage.”

Machine learning

Growing use of machine learning is exacerbating the problem. According to Beckwith, the utilization of machine learning has increased 40% a year since 2018 due to the higher volume and greater complexity of process data being generated.

And this has increased demand for staff with related expertise like support vector machine (SVM) technology he says. “SVMs are machine learning algorithms that can be applied in classification of microorganisms, protein sequence classification, predicting protein-protein interactions, as well as product quality and fault detection systems.”

Drug firms also struggle to attract staff who understand the infrastructure through which bioprocessing systems exchange data.

Beckwith says “AWS, the most widely used cloud platform, accounts for one of the largest skills gaps, with elevated demand also being seen for people with expertise in Microsoft Azure. Other systems with supply shortages include Jenkins and Docker.”

In addition, the industry finds it hard to hire programmers fluent in the languages used to control digital manufacturing technologies, according to Beckwith.

“Python is the most significant gap, despite it being one of the most available skills in the relevant talent market from other sectors. The difficulties in cross-fertilization from other industries are primarily due to compensation parity—in short, the other sectors are often willing to pay more.

“Other in-demand programming languages for extensive data analysis in software engineering include C++ and C. Big data processing and database management tools are the main aspects of data-related job requirements for data acquisition, processing, and storage.”


Addressing the data skills shortfall requires effort from every part of the biopharmaceutical ecosystem. Beckwith suggests, as a first step, drug manufacturers need to change how they think about employees.

“Companies should treat ‘talent’ as intangible intellectual assets. Talent is linked to business success. I have research illustrating a correlation between talent and financial ratios, which means talent management is analogous to IP and other assets.”

Companies should also consider using specialists to attract talent, Beckwith says, explaining, “Although such organizations are often perceived as expensive, research suggests internal recruiters with no experience are more costly.”

Technology suppliers could do more to ensure people working in the sector have the skills required according to Beckwith, who points to companies like Cytiva and MilliporeSigma as examples of organizations offering the most advanced programs.

Source link

News Science