Publicly available and validated DNA reference sequences useful for phylogeny estimation and identification of fungal pathogens are an increasingly important resource in the efforts of plant protection organizations to facilitate safe international trade of agricultural commodities. Colletotrichum species are among the most frequently encountered and regulated plant pathogens at U.S. ports-of-entry. The RefSeq Targeted Loci (RTL) project at NCBI (BioProject no. PRJNA177353) contains a database of curated fungal internal transcribed spacer (ITS) sequences that interact extensively with NCBI Taxonomy, resulting in verified name−strain−sequence type associations for >12,000 species. We present a publicly available dataset of verified and curated name−type strain−sequence associations for all available Colletotrichum species. This includes an updated GenBank Taxonomy for 238 species associated with up to 11 protein coding loci and an updated RTL ITS dataset for 226 species. We demonstrate that several marker loci are well suited for phylogenetic inference and identification. We improve understanding of phylogenetic relationships among verified species, verify or improve phylogenetic circumscriptions of 14 species complexes, and reveal that determining relationships among these major clades will require additional data. We present detailed comparisons between phylogenetic and similarity-based approaches to species identification, revealing complex patterns among single marker loci that often lead to misidentification when based on single-locus similarity approaches. We also demonstrate that species-level identification is elusive for a subset of samples regardless of analytical approach, which may be explained by novel species diversity in our dataset and incomplete lineage sorting and lack of accumulated synapomorphies at these loci.

The analysis of deletions may reveal evolutionary trends and provide new insight into the surprising variability and rapidly spreading capability that SARS-CoV-2 has shown since its emergence. To understand the factors governing genomic stability, it is important to define the molecular mechanisms of deletions in the viral genome. In this work, we performed a statistical analysis of deletions. Specifically, we analyzed correlations between deletions in the SARS-CoV-2 genome and repetitive elements and documented a significant association of deletions with runs of identical (poly-) nucleotides and direct repeats. Our analyses of deletions in the accessory genes of SARS-CoV-2 suggested that there may be a hypervariability in ORF7A and ORF8 that is not associated with repetitive elements. Such recurrent search in a “sequence space” of accessory genes (that might be driven by natural selection) did not yet cause increased viability of the SARS-CoV-2 variants. However, deletions in the accessory genes may ultimately produce new variants that are more successful compared to the viral strains with the conventional architecture of the SARS-CoV-2 accessory genes.

Shiga toxins in Shiga toxin-producing Escherichia coli (STEC) infections are responsible for bloody diarrhea and serious complications such as hemolytic uremic syndrome. Two types of toxins have been identified: Shiga toxin type 1 (Stx1) and the immunologically distinct Shiga toxin type 2 (Stx2). Numerous STEC that express toxin variants within those two major groups have been characterized, some of which confer unique biological properties. These variants are grouped within the Stx1 or Stx2 types and are often assigned subtypes to indicate they are not identical in sequence or phenotype. Because serious outcomes of infection are associated with certain Stx subtypes, there is a need to assign Stx sequences to the proper subtype. Here, we report a comprehensive analysis of known Stx subtypes and describe a scheme and algorithm to classify the Stx toxins and Stx operon sequences by phylogenetic sequence-based relatedness of the holotoxin conforming to historical type designations. We used this analysis to develop the free and open-source StxTyper software and database that implements this typing algorithm; StxTyper is also integrated into AMRFinderPlus 4.0 at the National Center for Biotechnology Information (NCBI). We validated and compared the results to PCR assays on a set of isolates and current state of the art surveillance methods used in the Danish public health system, and we summarize StxTyper results for over 111,000 publicly available E. coli genomes. We further propose a procedure to coordinate naming and identification for newly discovered and characterized Stx subtypes.