Skip to main content
Fig. 5 | BMC Bioinformatics

Fig. 5

From: NGS data vectorization, clustering, and finding key codons in SARS-CoV-2 variations

Fig. 5

Features that played a major role in the classification of each cluster extracted through Shapely Value (SHAP): Each graph shows which codon change had a great effect on distinguishing cluster 9 from other clusters, and from the top, they are listed in order of the most influential variable among 3,721 variables. In other graphs other than the first graph comparing clusters 9 and 1, the smaller the variable value is, the closer to the 9th cluster, and the larger the variable value, the closer to the other clusters. Conversely, in the first graph, the larger the variable value, the closer it is to cluster 9, and the smaller it is, the closer it is to cluster 1

Back to article page