Figure 3
From: SARS-CoV-2 host prediction based on virus-host genetic features
PCA reduction, all datasets. Dataset-1, training dataset; Dataset-2, testing; Dataset-3, SARS-CoV-2; Dataset-4, Bat Coronavirus, HCoV and Pangolin Coronavirus. Despite the Dataset-4 sequences being phylogenetically close to the Dataset-3 SARS-CoV-2 sequences15, we can notice that all of them do not cluster together when using RSCU as feature. Both Dataset-3 and 4 sequences were classified by our model as closer to bat coronaviruses than human coronaviruses.
