Machine learning prediction of hypertension integrating polygenic risk scores in inner Eurasian populations - Takeaways - MDSpire

Machine learning prediction of hypertension integrating polygenic risk scores in inner Eurasian populations

  • By

  • Vera Tsvetkova

  • Aleksandra Denisova

  • Saleem Mansour

  • Layal Shaheen

  • Iskandar Hweijeh

  • Leushin Artem

  • Travin Grigorii

  • Dilya Turkmenova

  • Liya Valieva

  • Anna Kim

  • Dmitrii Kharitonov

  • Anna Ilinskaya

  • Maria Poptsova

  • Valery Ilinsky

  • Alexander Rakitko

  • July 8, 2026

  • 0 min

Share

  • 1

    The study evaluated polygenic risk scores (PRS) for hypertension in a cohort of 175,704 individuals from inner Eurasian populations.

  • 2

    PRSs for systolic and diastolic blood pressure showed significant differences between the top and bottom deciles, with odds ratios of 6.20 and 6.71, respectively.

  • 3

    The neural network model integrating PRSs and questionnaire-based risk factors achieved a test ROC-AUC of 0.8245, indicating strong predictive performance for hypertension.

  • 4

    All evaluated PRSs were consistently associated with hypertension across various ancestry groups in Russia and neighboring regions.

  • 5

    The systolic blood pressure PRS demonstrated the most robust transferability, remaining informative in genetically diverse cohorts.

Original Source(s)

Related Content