Overview

The first half of our tutorial explores the striking parallels between the explosion of Natural Language Processing (NLP) and the current transformation in genomics. Just as the internet provided the massive corpus of text needed to train GPT models, the plummeting costs of DNA sequencing and the global response to the COVID-19 pandemic have created a "genomic internet".

Key Topics

• From Text to Genes: We will compare how NLP handles human languages with how Genomic Language Models (gLMs) interpret DNA. We’ll discuss the "Grammar of Life"; how sequences of nucleotides function like syntax and semantics.

• The Genomic Data Explosion: A look at the historical trajectory from the Human Genome Project to the present day, where DNA data is now measured in petabytes.

• Model Architectures: An in-depth look at the evolution of models.

  • • • Encoders: Understanding the context of specific genomic regions.

      • Decoders: Predicting the next DNA tokens.

      • State-of-art (Evo2): How the latest models utilize long-context windows to capture distal interactions in the genome.

• Healthcare Implications: How these models bridge the gap between "genomic big data" and clinical action, from identifying rare disease variants to designing targeted vaccines.