Improved annotation of enzyme encoding genes using deep learning

Supplementary datasets for the paper "Improved annotation of enzyme encoding genes using deep learning". The datasets include Supplementary Dataset 1 (Visualization of the latent representations of enzyme sequences in the Swiss-Prot database using TMAP), Supplementary Dataset 2 (Commonly highlighted motifs for each EC number using DeepEC v2 neural network), Supplementary Dataset 3 (Sequences for each of strain specific alleles), Supplementary Dataset 4 (EC number prediction results for the y-ome proteins), and Supplementary Dataset 5 (EC numbers of 128,100,490 protein sequences in 70,600 genomes in NCBI).