Big Data Approaches Still Rarely Used to Study Primary Sjogren’s Syndrome, Researchers Find

José Lopes, PhD avatar

by José Lopes, PhD |

Share this article:

Share article via email
big data

Use of big data machine learning to study autoimmune diseases is on the rise, but little focus has been devoted to primary Sjogren’s syndrome (pSS), a French study reports.

The study, “Pathogenesis-based treatments in primary Sjogren’s syndrome using artificial intelligence and advanced machine learning techniques: a systematic literature review,” appeared in the journal Human Vaccines & Therapeutics.

Treatment decisions for patients with pSS are based on the initial evaluation of symptoms and risk of death. Even though recent research has provided a deeper understanding, pSS is still regarded as a complex disease and used as a study case for the development of pathogenesis-based treatments, which, in contrast to gene-based approaches, are pathway-specific and thereby could have broader applicability.

Recent advances in molecular tools have enabled a substantial increase in the amount of data that can describe a single patient, including genetic expression, cellular profile, and detection of metabolites. Decreasing costs of these technologies has further aided their use in large patient groups, but extracting information from these large data sets is a complex task.

To overcome this challenge, scientists have been using new approaches based on artificial intelligence and machine learning, including the study of systemic autoimmune diseases such as Sjogren’s syndrome. This has led to a marked growth in the number of publications in scientific and medical journals.

Researchers in this study aimed to report on all published studies using big data analysis and machine learning to evaluate pathogenesis-based treatments for autoimmune diseases, focusing on pSS. They conducted a systematic literature review using a semiautomatic strategy with a program called BIBOT to retrieve all research published in the past decade.

In total, the team analyzed 1,017 publications. This represented 101 articles per year. Of note, the investigators found a significant increase in the number of studies per year from 2008 (74 articles) to 2017 (138), as well as increases in the size of the patient groups and in the use of terms associated with machine learning.

Large patient groups may imply increased use of big data approaches, the researchers said. They also mentioned that the growing use of machine learning may be due to its ability to detect indirect relations between components involved in the onset of complex disorders such as pSS and their treatment

However, only 12 articles focused on pSS and none on pathogenesis-based treatments. Instead, these studies used artificial intelligence to identify new biomarkers, the investigators found.

“To conclude, medicine progressively enters the era of big data analysis and artificial intelligence, but these approaches are not yet used to describe pSS-specific pathogenesis-based treatment,” they wrote.