Advancing Newborn Genetic Screening with Machine Learning
A groundbreaking study demonstrates how machine learning can standardize gene selection in newborn genetic screening, improving accuracy and public health outcomes.
Over the past decade, efforts to enhance newborn genetic screening have expanded significantly, driven by advancements in genomic technologies. The BabySeq Project, initiated more than ten years ago, pioneered the integration of genomic sequencing results into newborn care, providing valuable insights into genetic conditions from birth. Today, more than 30 international programs are exploring the use of genomic sequencing (NBSeq) to identify genetic disorders early in life. However, these programs often vary widely in the genes they select for screening, leading to inconsistencies that can impact clinical outcomes.
A recent study by researchers at Mass General Brigham introduces a novel, data-driven approach to refining gene selection in NBSeq programs using machine learning. Published in Genetics in Medicine, the research emphasizes the importance of choosing relevant genes based on scientific evidence, natural history of conditions, and treatment effectiveness. Co-senior author Nina Gold, MD, highlighted that thoughtful gene selection is crucial for maximizing the benefits of genomic screening in newborns.
The study analyzed 4,390 genes across 27 NBSeq initiatives, revealing that only about 1.7%—74 genes—were included in over 80% of these programs. Key factors influencing gene inclusion were their presence on the U.S. Recommended Uniform Screening Panel, robust data on natural disease progression, and strong evidence supporting treatment options. To help standardize gene selection, the team developed a machine learning model incorporating 13 predictive factors. This model accurately forecasts gene inclusion decisions among various programs and provides a ranked list that can adapt to new scientific evidence and regional health priorities.
This innovative approach aims to harmonize NBSeq programs worldwide, ensuring more consistent and informed decision-making. As Green from Mass General Brigham noted, such tools can improve the impact of genetic screening by aligning practices with the latest research and public health goals, ultimately benefiting early diagnosis and intervention in newborns.
For more details, see the full study by Thomas Minten et al. in Genetics in Medicine (2025). Source: https://medicalxpress.com/news/2025-05-machine-newborn-genetic-screening.html
Stay Updated with Mia's Feed
Get the latest health & wellness insights delivered straight to your inbox.
Related Articles
Long-Term Cancer Control Strategies Show Significant Impact in Australia
Decades of persistent cancer control initiatives in Australia have prevented over 230,000 deaths, demonstrating the powerful impact of long-term prevention and early detection strategies.
Research Reveals US Self-Reported Race and Ethnicity Don't Accurately Reflect Genetic Ancestry
New research shows that self-reported race and ethnicity in the US do not accurately reflect an individual's true genetic ancestry, highlighting implications for personalized medicine.
How a Past Injury Can Lead to Lasting Changes in Stress and Pain Responses
Learn how past injuries can cause long-lasting changes in the brain, leading to heightened stress and pain responses, even after healing. New research reveals the neural mechanisms behind this phenomenon and potential therapeutic targets.
New Insights into How Viral Infections May Trigger Arthritis-Like Diseases
New research sheds light on how mosquito-borne viruses like Chikungunya can lead to long-term autoimmune-like joint pain, highlighting immune system mechanisms behind chronic inflammation.