Automatic Identification of SARS Coronavirus using Compression-Complexity Measures

Karthi Balasubramanian; Nithin Nagaraj

0.1648 ± 0.014 0.3801 ± 0.033 SARS-CoV-2 and Urbani 0.1625 ± 0.015 0.3749 ± 0.038 BJ01 and Urbani 0.1608 ± 0.014 0.3661 * ± 0.032 * indicates statistical significance at p < 0.05. Table 5 . It was found that only ETC based distance measure yielded the least value (statistically significant) for the SARS-CoV-1 virus pair (BJ01 and Urbani), and not the LZ based distance measure. The statistical validation of the results is depicted using 95% confidence interval plots as shown in Figure 4 . Based on the sample data, at an overall error rate of 5%, we can conclude that ETC is able to distinguish between SARS-CoV-1 and SARS-CoV-2 viruses by using only short contiguous segments consisting of 25 nucleic bases chosen independently from random locations (300 such segments) of the entire sequence. LZ based distance measure fails to do so.