From: Unsupervised clustering of SARS-CoV-2 using deep convolutional autoencoder
Ref | Objective | DNN | Input | Output | Accuracy |
---|---|---|---|---|---|
[27] | Viral classification | CNN + FC | The whole genomic sequences of a virus | Different viral classes | 96.7% (with noise) 98.7% (without noise) |
[28] | Viral host prediction | RC-CNN and RC-LSTM | Contigs of a genome | Human or nonhuman host | 91.7% CNN 86.3% LSTM |
[29] | Viral classification | Stacked sparse autoencoder (SSAE) | Image representations of the complete genome sequences | Different classes | 98.9% and 100% |
[30] | Predicting the mutation rate of SARS-CoV-2 | LSTM-RNN | Complete genome | Mutation rate calculations | (RMSE) of 0.06 in testing and 0.04 in training |
[31] | Predicting the similarity score of the genome of “SARS-CoV-2” with other viruses | CNN + LSTM | Genome sequence | Similarity score with other viruses | 99.27% |