Fig. 2
From: KEGG orthology prediction of bacterial proteins using natural language processing
Distribution of structural similarity metric TM-score in unmatch and added cases. These two cases represent instances where our pipeline incorrectly assigned the K number, while the KEGG GENES database assigned a different K number (unmatch) or did not assign K number (added). A TM-score of \(\ge 0.5\) suggests the presence of similar structural domains, while a TM-score of \(\ge 0.8\) indicates highly similar structures, which implies potential functional similarity