Fig. 1From: DeepECA: an end-to-end learning framework for protein contact prediction from a multiple sequence alignmentSchematic representation of weighted MSA: The left panel shows a part of the MSA. The right panel shows weight values for each sequence in the MSABack to article page