Hierarchical sets: analyzing pangenome structure through scalable set visualizations

Structure Learning for Gene Regulatory Networks

Inference of biological network structures is often performed on high-dimensional data, yet is hindered by the limited sample size of high throughput "omics" data typically available. To overcome this challenge, often referred to as the "small n, large p problem," we exploit known organizing principles of biological networks that are sparse, modular, and likely share a large portion of their underlying architecture. We present SHINE-Structure Learning for Hierarchical Networks-a framework for defining data-driven structural constraints and incorporating a shared learning paradigm for efficiently learning multiple Markov networks from high-dimensional data at large p/n ratios not previously feasible. We evaluated SHINE on Pan-Cancer data comprising 23 tumor types, and found that learned tumor-specific networks exhibit expected graph properties of real biological networks, recapture previously validated interactions, and recapitulate findings in literature. Application of SHINE to the analysis of subtype-specific breast cancer networks identified key genes and biological processes for tumor maintenance and survival as well as potential therapeutic targets for modulating known breast cancer disease genes.

Avian strains of emerging pathogen Escherichia fergusonii are phylogenetically diverse and harbor the greatest AMR dissemination potential among different sources: Comparative genomic evidence

Introduction

Escherichia fergusonii is regarded as an emerging pathogen with zoonotic potential. In the current study, we undertook source-wise comparative genomic analyses (resistome, virulome, mobilome and pangenome) to understand the antimicrobial resistance, virulence, mobile genetic elements and phylogenetic diversity of E. fergusonii.

Methods

Six E. fergusonii strains (5 multidrug resistant strains and 1 biofilm former) were isolated from poultry (duck faeces and retail chicken samples). Following confirmation by phenotypic and molecular methods, the isolates were further characterized and their genomes were sequenced. Comparative resisto-virulo-mobilome analyses and pangenomics were performed for E. fergusonii genomes, while including 125 other E. fergusonii genomes available from NCBI database.

Results and discussion

Avian and porcine strains of E. fergusonii were found to carry significantly higher number of antimicrobial resistance genes (p < 0.05) and mobile genetic elements (plasmids, transposons and integrons) (p < 0.05), while the pathogenic potential of bovine strains was significantly higher compared to other strains (p < 0.05). Pan-genome development trends indicated open pan-genome for all strains (0 < γ < 1). Genomic diversity of avian strains was found to be greater than that from other sources. Phylogenetic analysis revealed close clustering among isolates of similar isolation source and geographical location. Indian isolates of E. fergusonii clustered closely with those from Chinese and a singleton Australian isolate. Overall, being the first pangenomic study on E. fergusonii, our analysis provided important cues on genomic features of the emerging pathogen E. fergusonii while highlighting the potential role of avian strains in dissemination of AMR.

The pangenome of the wheat pathogen Pyrenophora tritici-repentis reveals novel transposons associated with necrotrophic effectors ToxA and ToxB

BACKGROUND

In fungal plant pathogens, genome rearrangements followed by selection pressure for adaptive traits have facilitated the co-evolutionary arms race between hosts and their pathogens. Pyrenophora tritici-repentis (Ptr) has emerged recently as a foliar pathogen of wheat worldwide and its populations consist of isolates that vary in their ability to produce combinations of different necrotrophic effectors. These effectors play vital roles in disease development. Here, we sequenced the genomes of a global collection (40 isolates) of Ptr to gain insights into its gene content and genome rearrangements.

RESULTS

A comparative genome analysis revealed an open pangenome, with an abundance of accessory genes (~ 57%) reflecting Ptr's adaptability. A clear distinction between pathogenic and non-pathogenic genomes was observed in size, gene content, and phylogenetic relatedness. Chromosomal rearrangements and structural organization, specifically around effector coding genes, were detailed using long-read assemblies (PacBio RS II) generated in this work in addition to previously assembled genomes. We also discovered the involvement of large mobile elements associated with Ptr's effectors: ToxA, the gene encoding for the necrosis effector, was found as a single copy within a 143-kb 'Starship' transposon (dubbed 'Horizon') with a clearly defined target site and target site duplications. 'Horizon' was located on different chromosomes in different isolates, indicating mobility, and the previously described ToxhAT transposon (responsible for horizontal transfer of ToxA) was nested within this newly identified Starship. Additionally, ToxB, the gene encoding the chlorosis effector, was clustered as three copies on a 294-kb element, which is likely a different putative 'Starship' (dubbed 'Icarus') in a ToxB-producing isolate. ToxB and its putative transposon were missing from the ToxB non-coding reference isolate, but the homolog toxb and 'Icarus' were both present in a different non-coding isolate. This suggests that ToxB may have been mobile at some point during the evolution of the Ptr genome which is contradictory to the current assumption of ToxB vertical inheritance. Finally, the genome architecture of Ptr was defined as 'one-compartment' based on calculated gene distances and evolutionary rates.

CONCLUSIONS

These findings together reflect on the highly plastic nature of the Ptr genome which has likely helped to drive its worldwide adaptation and has illuminated the involvement of giant transposons in facilitating the evolution of virulence in Ptr.

An Acinetobacter non-baumannii Population Study: Antimicrobial Resistance Genes (ARGs)

Acinetobacter non-baumannii species are becoming common etiologic agents of nosocomial infections. Furthermore, clinical isolates belonging to this group of bacteria are usually resistant to one or more antibiotics. The current information about antibiotic resistance genes in the different A. non-baumannii species has not yet been studied as a whole. Therefore, we did a comparative study of the resistomes of A. non-baumannii pathogens based on information available in published articles and genome sequences. We searched the available literature and sequences deposited in GenBank to identify the resistance gene content of A. calcoaceticus, A. lwoffii, A. junii, A. soli, A. ursingii, A. bereziniae, A. nosocomialis, A. portensis, A. guerrae, A. baylyi, A. calcoaceticus, A. disperses, A. johnsonii, A. junii, A. lwoffii, A. nosocomialis, A. oleivorans, A. oryzae, A. pittii, A. radioresistens, and A. venetianus. The most common genes were those coding for different β-lactamases, including the carbapenemase genes bla_NDM-1 and bla_OXA-58. A. pittii was the species with the most β-lactamase resistance genes reported. Other genes that were commonly found include those encoding some aminoglycoside modifying enzymes, the most common being aph(6)-Id, ant(3″)-IIa, and aph(3″)-Ib, and efflux pumps. All or part of the genes coding for the AdeABC, AdeFGH, and AdeIJK efflux pumps were the most commonly found. This article incorporates all the current information about A. non-baumannii resistance genes. The comparison of the different resistomes shows that there are similarities in the genes present, but there are also significant differences that could impact the efficiency of treatments depending on the etiologic agent. This article is a comprehensive resource about A. non-baumannii resistomes.

Use of Genomics to Investigate Historical Importation of Shiga Toxin-Producing Escherichia coli Serogroup O26 and Nontoxigenic Variants into New Zealand

Shiga toxin-producing Escherichia coli serogroup O26 is an important public health pathogen. Phylogenetic bacterial lineages in a country can be associated with the level and timing of international imports of live cattle, the main reservoir. We sequenced the genomes of 152 E. coli O26 isolates from New Zealand and compared them with 252 E. coli O26 genomes from 14 other countries. Gene variation among isolates from humans, animals, and food was strongly associated with country of origin and stx toxin profile but not isolation source. Time of origin estimates indicate serogroup O26 sequence type 21 was introduced at least 3 times into New Zealand from the 1920s to the 1980s, whereas nonvirulent O26 sequence type 29 strains were introduced during the early 2000s. New Zealand's remarkably fewer introductions of Shiga toxin-producing Escherichia coli O26 compared with other countries (such as Japan) might be related to patterns of trade in live cattle.

Updating the genomic taxonomy and epidemiology of Campylobacter hyointestinalis

Campylobacter hyointestinalis is a member of an emerging group of zoonotic Campylobacter spp. that are increasingly identified in both gastric and non-gastric disease in humans. Here, we discovered C. hyointestinalis in three separate classes of New Zealand ruminant livestock; cattle, sheep and deer. To investigate the relevance of these findings we performed a systematic literature review on global C. hyointestinalis epidemiology and used comparative genomics to better understand and classify members of the species. We found that C. hyointestinalis subspecies hyointestinalis has an open pangenome, with accessory gene contents involved in many essential processes such as metabolism, virulence and defence. We observed that horizontal gene transfer is likely to have played an overwhelming role in species diversification, favouring a public-goods-like mechanism of gene ‘acquisition and resampling’ over a tree-of-life-like vertical inheritance model of evolution. As a result, simplistic gene-based inferences of taxonomy by similarity are likely to be misleading. Such genomic plasticity will also mean that local evolutionary histories likely influence key species characteristics, such as host-association and virulence. This may help explain geographical differences in reported C. hyointestinalis epidemiology and limits what characteristics may be generalised, requiring further genomic studies of C. hyointestinalis in areas where it causes disease.