Banca de DEFESA: DANIELE ALMEIDA ALVES

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
DISCENTE : DANIELE ALMEIDA ALVES
DATA : 08/10/2020
HORA: 13:30
LOCAL: Plataforma Conferência Web da RNp
TÍTULO:

Evaluation of the use of global genomic similarity indices for classification of emerging pathogenic species of the gender Corynebacterium

PALAVRAS-CHAVES:

Bacterial genomics, Average identity of nucleotides, tetranucleotides, genomic taxonomy, Corynebacterium diphtheriae, Emerging pathogens; Corynebacterium spp

PÁGINAS: 68
GRANDE ÁREA: Outra
ÁREA: Tecnologia e Inovação
RESUMO:

BACKGROUND: Overall Genome-Relatedness Indexes (OGRIs) have been extensively used in recent years for taxonomic classification of bacteria. Average Nucleotide Identity by BLAST (ANIb) in pairwise genome comparisons is widely regarded as the most accurate index for bacterial species circumscription, when considering a species boundary of ca. 95- 96% identity. However, the sole use of ANIb for species identification may render confusing results for some closely related bacterial groups, such as many pathogenic Corynebacterium spp. AIM: To evaluate the performance of an in-house developed speciesclassifier, based on the correlation values between different genome relatedness indexes, to correctly classify genomic sequences from the Corynebacterium diphtheriae group. METHODS: 213 genomic sequences corresponding to three closely related Corynebacterium species were retrieved from NCBI’s Genome DB: 188 classified as C. diphtheriae and 03 isolates of C. diphtheriae subsp. Lausannense; 10 as Corynebacterium belfantii; 01 as Corynebacterium rouxii and 11 classified as C. diphtheriae obtained from European Nucleotide Archive. Tetranucleotide usage patterns (TETRA) and ANIb were calculated through the JSpecies Web server application and compared all-vs-all. Resulting matrices were merged to generate a single fingerprint matrix, which was used to compute Spearman’s correlation values among bacterial genomes using an in-house script developed on R software. Multilocus sequence analysis (MLSA) with six house-keeping genes and split-decomposition analyses were used to confirm relationships between the various species. RESULTS: In total, 45,369 genome-to-genome comparisons composed the fingerprint matrix that was used to build a dendrogram with welldefined clades (>95% bootstrap confidence). The groups containing C. belfantii and C. rouxii were clearly distinguished by this strategy, as opposed to the use of ANIb alone. Additionally, we observed that our results are corroborated with the MLSA, highlighting the mistaken classification in the NCBI. We also observed that Lausannense subspecies form demarcated clades of diphtheria. CONCLUSIONS: An in-house developed classifier that integrates different OGRIs results was the most efficient tool for species circumscription in the C. diphtheriae group, when compared to ANIb alone. We anticipate that this new strategy can be extrapolated to improve genome-based identification of other clinically important bacterial pathogens.

MEMBROS DA BANCA:
Externo à Instituição - Aristóteles Góes-Neto
Presidente - 1621863 - ERIC ROBERTO GUIMARAES ROCHA AGUIAR
Externo ao Programa - 2415395 - THIAGO LUIZ DE PAULA CASTRO