Fig. 1
From: High-throughput deep learning variant effect prediction with Sequence UNET
Model overview. A Sequence UNET model schematic. Blue rectangles represent intermediate layer output matrices and green the final prediction. B Bar chart showing the computation time taken to compute predictions for all variants in SARS-CoV-2 Spike protein by two commonly used tools, SIFT4G and FoldX, and Sequence UNET. These tools were chosen for a previous analysis, but broadly span the typical timescales of current tools