Nilah M. Ioannidis

The Enformer (Avsec et al., 2021) is a sequence-based prediction model, which leverages the attention mechanism in a transformer neural network to capture long-range effects on gene regulation. For Enformer predictions, we subset the 5313 ChIP-seq, DNase-seq, ATAC-seq, and CAGE tracks to only those relevant to the lymphoblastoid cell line (GM12878), the cell line with respect to which GEUVADIS gene expression phenotypes are measured.

Abstract Deep learning models in genomics that predict molecular phenotypes from DNA sequence traditionally focus on one-hot encoded representations. Here, we develop a novel model that extends this approach by incorporating DNA structural attributes indicative of local shape alongside canonical sequence inputs. This augmentation provides an additional axis for model interpretability and aids in identifying regulatory patterns not apparent from sequence alone.

Abstract The nascent field of genomic AI is rapidly expanding with new models, benchmarks, and findings. As the field diversifies, there is an increased need for a common set of measurement tools and perspectives to standardize model evaluation. Here, we present a statistically grounded framework for performance evaluation, visualization, and interpretation using the prominent sequence-based deep learning models Enformer and Borzoi as case studies.