Neural network based integration of assays to assess pathogenic potential

Limited data significantly hinders our capability of biothreat assessment of novel bacterial strains. Integration of data from additional sources that can provide context about the strain can address this challenge. Datasets from different sources, however, are generated with a specific objective and which makes integration challenging. Here, we developed a deep learning-based approach called the neural network embedding model (NNEM) that integrates data from conventional assays designed to classify species with new assays that interrogate hallmarks of pathogenicity for biothreat assessment. We used a dataset of metabolic characteristics from a de-identified set of known bacterial strains that the Special Bacteriology Reference Laboratory (SBRL) of the Centers for Disease Control and Prevention (CDC) has curated for use in species identification. The NNEM transformed results from SBRL assays into vectors to supplement unrelated pathogenicity assays from de-identified microbes. The enrichment resulted in a significant improvement in accuracy of 9% for biothreat. Importantly, the dataset used in our analysis is large, but noisy. Therefore, the performance of our system is expected to improve as additional types of pathogenicity assays are developed and deployed. The proposed NNEM strategy thus provides a generalizable framework for enrichment of datasets with previously collected assays indicative of species.

A statistical approach to identify prevalent virulence factors responsible for post-weaning diarrhoeic piglets

A statistical approach was carried out to identify the prevalent virulence factors responsible for post-weaning diarrhoea (PWD). Healthy piglets' faecal samples and diarrhoeic piglets' rectal swab specimens were secured. Twenty-six (26) and 100 independent enterotoxigenic Escherichia coli (ETEC) strains were subsequently isolated. These strains were assessed utilising polymerase chain reaction to identify the encoding genes of six virulence factors: heat-labile enterotoxin (LT; encoded by eltAB), heat-stable enterotoxin A (STa; encoded by estA), heat-stable enterotoxin B (STb; encoded by estB), enteroaggregative E. coli heat-stable enterotoxin 1 (EAST1; encoded by astA), Shiga toxin 2e (Stx2e; encoded by stx2e), and F18 fimbriae (encoded by fedA). The LT and ST secretions were investigated using enzyme-linked immunosorbent assays. From direct observation, no stx2e was evident in the 126 strains. Among the 26 strains retrieved from the healthy piglets, none harboured fedA or secreted LT; 23% (6/26) secreted ST, and 50% (13/26) carried astA. A statistical regression was applied on the 100 E. coli strains retrieved from the diarrhoeic piglets, where fedA was set as the dependent variable and the enterotoxin secretions were set as the independent variables. The results exhibit that the LT secretion was the only significant factor (P < 0.000 1) correlated to fedA in the diarrhoeic piglets; thus, it is concluded that the prevalent virulence factors for PWD were the ECET strain with F18 fimbriae adhesion and LT secretion, but not astA or stx2e.

Pedobacter fastidiosus sp. nov., isolated from glacial habitats of maritime Antarctica

Strains P8930T and 478 were isolated from Antarctic glaciers located on James Ross Island and King George Island, respectively. They comprised Gram-stain-negative short rod-shaped cells forming pink pigmented colonies and exhibited identical 16S rRNA gene sequences and highly similar MALDI TOF mass spectra, and hence were assigned as representatives of the same species. Phylogenetic analysis based on 16S rRNA gene sequences assigned both isolates to the genus Pedobacter and showed Pedobacter frigidisoli and Pedobacter terrae to be their closest phylogenetic neighbours, with 97.4 and 97.2 % 16S rRNA gene sequence similarities, respectively. These low similarity values were below the threshold similarity value of 98.7%, confirming the delineation of a new bacterial species. Further genomic characterization included whole-genome sequencing accompanied by average nucleotide identity (ANI) and digital DNA–DNA hybridization calculations, and characterization of the genome features. The ANI values between P8930T and P. frigidisoli RP-3-11T and P. terrae DSM 17933T were 79.7 and 77.6 %, respectively, and the value between P. frigidisoli RP-3-11T and P. terrae DSM 17933T was 77.7 %, clearly demonstrating the phylogenetic distance and the novelty of strain P8930T. Further characterization included analysis of cellular fatty acids, quinones and polar lipids, and comprehensive biotyping. All the obtained results proved the separation of strains P8930T and 478 from the other validly named Pedobacter species, and confirmed that they represent a new species for which the name Pedobacter fastidiosus sp. nov. is proposed. The type strain is P8930T (=CCM 8938T=LMG 32098T).