Decoding the Human Genome with Deep Learning
summary
Machine learning builds a predictive regulatory map across the 3 billion letters of the human genome so single‑letter variants get biochemical consequences instead of vague association signals. A convolutional deep learning classifier (DeepSEA model) learns patterns of transcription‑factor binding and chromatin signals and predicts when an A→C disrupts binding, lowers expression, and plausibly predisposes to autism. This targets the vast noncoding space — >98% regulatory — where traditional population studies lack power.
AI-written summary · about
related talks
38mInterpreting Deep Learning Models Of Functional Genomics Data To Decode Regulatory Sequence...
01h 13mIntroduction to Regulatory Genomics and Epigenomics I: Intro to the Biology of Gene Regulation
30mMachine Learning: A New Approach to Drug Discovery
43mAlgorithms for Inferring Phenotypes from Ancient DNA
12mHow data science can make your genome actionable
37mDynamics of 3D Enhancer-Gene Associations Across Diverse Human Cell Types and Tissues
01h 19mAnalyzing Genetic Networks Using Neural Networks
27mPredicting the Deleteriousness of Genomic Variants – Big and Small