RNA genome conservation and secondary structure in SARS-CoV-2 and SARS-related viruses

Ramya Rangan; Ivan N. Zheludev; Rhiju Das

1. The first multiple sequence alignment (SARSr-MSA-1) was computed by aligning sequences curated by Ceraolo and Giorgi, 10 filtered by including only the reference genome sequence NC_0405512.2 11 from the SARS-CoV-2 sequence set, removing the two MERS sequences, and leaving in all remaining betacoronavirus whole genome sequences. This alignment captures a range of SARS-related bat coronavirus and SARS sequences with only 11 sequences. These sequences correspond well to the SARS-related group defined in Gorbalenya, Baker, et al. 12 2. The second MSA (SARSr-MSA-2) was obtained from BLAST by searching for the 100 complete genome sequences closest to the SARS-CoV-2 reference genome. This alignment captures a larger set of SARS-CoV-2, SARS, and bat coronavirus sequences than SARSr-MSA-1 but includes many sequences with high pairwise similarity. 3. The final MSA (SARSr-MSA-3) was obtained by locating all complete genome betacoronavirus sequences from the NCBI database, and removing mutually similar sequences with a 99% sequence conservation cutoff. With 180 sequences with at most 99% pairwise sequence similarity, this MSA captures a broader set of betacoronaviruses than SARSr-MSA-1 and SARSr-MSA-2 but is more challenging to align due to higher sequence diversity.