This podcast explores a new study published in the European Radiology journal examining the progression of arterial calcification over a 14 year period. We'll uncover what causes these 'hidden harbingers' of disease, learn where they occur in our bodies, and discover how cardiovascular risk factors and sex impact their development. Join us as we translate the science and explain why understanding arterial calcification matters for your health

In this series, we dive into the fascinating world of genetics and how our DNA can help us understand the causes of diseases. This episode focuses on a groundbreaking technique called Mendelian Randomization, which uses genetic differences to figure out how certain risk factors—like high blood pressure or diabetes—affect our health.

The catch? Many of these factors are either "yes" or "no" (you either have the condition, or you don’t), but the science behind them is much more complex. We explore how researchers deal with these challenges, using our genes to make sense of the messy reality behind conditions like hypertension. Learn how scientists use genetics to uncover potential ways to prevent diseases and improve health outcomes, all while facing the tricky task of dealing with binary choices in data.

Whether you're into science or just curious about how your genes influence your health, this episode will give you a clearer picture of how researchers use Mendelian Randomization to unlock the secrets of our genetic code.

Based on the paper:

Have you ever wondered how scientists are using cutting-edge technology to understand our DNA?

Join us as we explore the fascinating world of deep learning in genomics and discuss a new paper that examines how to make these powerful tools more interpretable.

Deep learning, a type of artificial intelligence, is revolutionizing many fields, including genomics. By analyzing vast amounts of genetic data, these complex algorithms can uncover hidden patterns and make predictions about our health, traits, and even our ancestry.

But there's a catch: deep learning models are often described as "black boxes" because it's difficult to understand how they reach their conclusions. This lack of transparency can make it challenging for scientists to trust and fully utilize these powerful tools. This is where interpretable deep learning comes in.

The paper we'll be discussing, "Designing Interpretable Deep Learning Applications in Genomics: a Quantitative Analysis", explores different approaches to open up the black box and shed light on the decision-making processes of these complex models

The Rotterdam Study is a long-term, population-based cohort study in the Netherlands focused on identifying risk factors and understanding the progression of multifactorial diseases, particularly in mid-life and older individuals. The research encompasses a broad range of disease areas including cardiovascular disease, diabetes, internal medicine, pulmonary disease, dermatology, neurological diseases, mental health, musculoskeletal diseases, otolaryngology, liver disease, population imaging, and population genomics. Researchers at the Rotterdam Study utilise various data collection methods including questionnaires, physical examinations, biological samples, and automated linkage to health records. They are actively investigating the roles of genetics, lifestyle, and environmental factors on disease development and progression.