Fig. 6From: Improving protein domain classification for third-generation sequencing reads using deep learningThe mean, min, and max value of classification results of ProDOMA, HMMER, and DeepFam on classifying four sets of simulated long reads with different error rates. Each test set contains roughly 228K simulated reads from 86 sub-subfamilies. X-axis: different error rates. Y-axis: classification accuracyBack to article page