Table 2 MeRG fit parameters.

From: Variant-driven early warning via unsupervised machine learning analysis of spike protein mutations for COVID-19

Region

Standard variant

Variant of concern

Transmissibility

VoC percentage

A

\(\gamma\)

\(A_{\mathrm{VoC}}\)

\(\gamma _{\mathrm{VoC}}\)

VoC/VoI

Increase

\(A_{\%}\)

\(\gamma _{\%}\)

UK

2140 (12)

0.0668 (5)

2530 (10)

0.0994 (7)

Alpha

49%

97.3 (3)%

0.076 (1)

  

–

–

Delta

–

99 (1)%

0.115 (2)

South Africa

1104 (2)

0.0705 (4)

1161 (2)

0.0904 (5)

Beta

28%

91.9 (8)%

0.061 (4)

  

–

–

Delta

–

96 (6)%

0.090 (7)

India

717 (3)

0.0358 (4)

497.8 (8)

0.0858 (3)

Alpha

140%

–

–

  

908 (5)

0.0747 (6)

Delta

109%

–

–

California

4773 (7)

0.0620 (3)

2250 (5)

0.0758 (5)

Epsilon

22%

59.9 (6)%

0.059 (4)

  

–

–

Alpha

–

61.0 (6)%

0.0610 (2)

  1. Parameters from the fit of the VoC/VoI for the UK, South Africa, California and India, also shown in Fig. 6. The fit follows the MeRG model, according to which each variant can be fitted by an independent logistic function. For the UK, the “standard variant” fit corresponds to the first peak, in October–November 2020. The transmissibility increase is computed by comparing the gamma of the VoC with that of the standard variant in the same country. For the new variants that have not reached the peak of diffusion, it is not possible to extract reliable values for the eRG parameters.
Back to article page