Intra-Host Co-Existing Strains of SARS-CoV-2 Reference Genome Uncovered by Exhaustive Computational Search

Figures

• 

  Figure 1. Strain identification process. The process begins with extracting non-human reads from the sample and applying double-model error correction to it. Aligning the corrected reads with the SARS-CoV-2 reference genome would allow the selection of reads that are related to the virus. From the related reads, select those with mismatches to the reference genome as candidate reads possibly belonging to co-existing strains. Then, proceed with the substitution process for identifying the co-existing strains from the reads that are compatible with each other.

• 

  Figure 2. The coverage at each position of the sequences. The coverage of a base is calculated by how many reads cover it. The average sequence coverage for the reference genome was higher than both discovered strains. At some positions, the discovered strains’ sequences had a higher coverage than the reference genome sequence, which indicates identification of more dominant bases in the discovered strains.

• 

  Figure 3. Phylogenetic tree constructed from nucleotide sequences using FastTree. The number denotes the confidence of the maximum-likelihood calculation by FastTree. It shows with a high confidence that the two discovered strains are closely related to the reference genome.

• 

  Figure 4. Phylogenetic tree constructed from the spike protein sequences. The two zero edges are persistent with both Fasttree [26] and iQTree2 [38].

• 

  Figure 5. Phylogenetic tree constructed from the nucleotide sequences using PHYLIP’s dnapars, a parsimony-based software for phylogenetic tree construction. Dnapars does not provide a confidence value for the constructed tree, which is different from Fasttree.

• 

  Figure 6. Phylogenetic tree constructed from nucleotide sequences of MR discovered strains.

• 

  Figure 7. Phylogenetic tree constructed from the spike protein sequences of MR discovered strains.