Mutational dynamics of the SARS coronavirus in cell culture and human populations isolated in 2003

Vega, Vinsensius B; Ruan, Yijun; Liu, Jianjun; Lee, Wah Heng; Wei, Chia Lin; Se-Thoe, Su Yun; Tang, Kin Fai; Zhang, Tao; Kolatkar, Prasanna R; Ooi, Eng Eong; Ling, Ai Ee; Stanton, Lawrence W; Long, Philip M; Liu, Edison T

We tested the biological validity of this approach by examining the mutational frequency of known genes in the SARS genome. Because of the importance of the 3CL protease and the polymerase for viral replication, we sus-pected that true non-synonymous mutations in the SARS-CoV present in clinical samples might be rare in these two ORFs in comparison to other structural genes such as those encoding the S, M, and N proteins. Without a mutation filter, sequence variations are commonly observed in the 3CL protease and the polymerase genes. However, when mutations are identified only as variants seen in two or more isolates, then no mutations are detected in the critical 3CL protease and polymerase genes, whereas mutations are noted in the S, M, and N genes regardless of the filter stringency ( Figure 2 ). Therefore, we determined that the most effective mutation filter is presence of a sequence variant in more than two isolates.