Table 1 Average of performance metrics for each class in 100 independent simulated datasets.
From: PACIFIC: a lightweight deep-learning classifier of SARS-CoV-2 and co-infecting RNA viruses
Class | FNR (± 95% CI) | FPR (± 95% CI) | Precision (± 95% CI) | Recall (± 95% CI) | Accuracy (± 95% CI) |
|---|---|---|---|---|---|
Coronaviridae | 0.116% (2.14e−05) | 0.008% (2.38e−06) | > 0.999 (1.41e−05) | 0.999 (2.14e−05) | > 0.999 (3.67e−06) |
Influenza | 0.153% (2.65e−05) | < 0.001% (3.18e−07) | > 0.999 (1.95e−06) | 0.999 (2.65e−05) | > 0.999 (3.80e−06) |
Metapneumovirus | 0.340% (3.62e−05) | < 0.001% (1.04e−07) | > 0.999 (6.23e−07) | 0.997 (3.62e−05) | > 0.999 (5.20e−06) |
Rhinovirus | 0.121% (1.90e−05) | 0.004% (1.41e−06) | > 0.999 (8.44e−06) | 0.999 (1.90e−05) | > 0.999 (2.79e−06) |
SARS-CoV-2 | 0.222% (3.27e−05) | < 0.001% (1.43e−07) | > 0.999 (8.63e−07) | 0.998 (3.27e−05) | > 0.999 (4.64e−06) |
Human | 0.037% (1.32e−05) | 0.166% (1.03e−05) | 0.501 (1.45e−05) | > 0.999 (1.32e−05) | 0.858 (8.94e−06) |
