While other researchers are focusing on population-level risk assessment, scientists at the Max Planck Institute for Molecular Genetics are addressing a different challenge: interpreting the biological activity encoded in DNA sequences.
Unlike earlier sequence-based models, the AI tool Corgi (Context-aware Regulatory Genomics Inference) incorporates information about regulatory gene expression, allowing predictions that account for the cellular environment in which genes operate. The researchers report that the model can predict multiple indicators of genome activity, including gene expression and chromatin accessibility.
Although the work is primarily intended for basic research, it addresses a challenge that extends beyond genomics: deriving biologically relevant insights from large-scale molecular data. Improved interpretation of disease-relevant variants could strengthen the analytical foundation for precision diagnostics and future research into disease mechanisms.