The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Deep Learning for Population Genetic Inference with Dan Schrider - TWiML Talk #249
Apr 9, 2019
In this discussion, Dan Schrider, an assistant professor at UNC Chapel Hill, reveals how machine learning is revolutionizing population genetics. He highlights the effectiveness of convolutional neural networks in outperforming traditional statistical methods. Dan shares insights on the integration of bioinformatics and discusses the importance of reference genomes in understanding genetic variation. Furthermore, he explores the challenges of estimating human population dynamics and the need for innovative approaches to bridge machine learning with genetic research.
AI Snips
Chapters
Transcript
Episode notes
Population Genetics Overview
- Population genetics studies the evolutionary dynamics of gene sequences, especially recent evolutionary history.
- Machine learning helps analyze high-dimensional genomic data, uncovering evolutionary forces like natural selection.
Recent Evolutionary Events
- Population genetics focuses on recent evolutionary events, within hundreds of thousands of years for humans.
- This timeframe offers better resolution for studying variations and forces shaping present-day genetic patterns.
Genome Sample Analysis
- Genome samples are analyzed as matrices, with rows representing genomes and columns representing sites.
- Variations at these sites provide insights into evolutionary forces shaping genetic diversity.