Biology subjects

Ashkenazy, H.

Publications and source records attributed to Ashkenazy, H..

11 recordsLinked to original sources

bioRxiv2026

Dynamic co-existence of bacteriophages and their hosts in the Arabidopsis thaliana phyllosphere

Bacterial communities and the bacteriophages infecting them are the basis of every ecosystem, including holobionts. The various ways in which these microorganisms interact with each other in complex communities over the life of the host affects the holobiont fitness. Despite being ubiquitous and environmentally relevant, plant-associated microbial communities remain understudied, especially in the phyllosphere, mainly because of the low abundance of microbes and the complexity of the system. In this work we followed bacteria and phage community dynamics in the phyllosphere over a growing cycle of Arabidopsis thaliana, to understand the ecology and relevance of bacteriophages in complex bacterial communities. We focused on Pseudomonas, a common plant pathogen and commensal, and the phages infecting them, in three setups of increasing complexity: in vitro, controlled experiments in planta and in wild populations of A. thaliana. We found that bacterial communities are resilient to phage infection, and more dynamic than the phages infecting them over the growing season, suggesting that although ubiquitous and abundant, bacteriophages exert selective pressures on leaf bacterial communities only intermittently.

microbiology↗

bioRxiv2026

Biosynthetic gene clusters in Pseudomonas viridiflava have a fitness cost during Arabidopsis thaliana infection

Specialized or secondary metabolites mediate biotic interactions, including virulence and defense. In plant-pathogenic Pseudomonas, certain specialized metabolites can enhance colonization of plant hosts, yet their broader contribution to plant-microbe interactions and the relative importance of different metabolites remain unclear. Specialized metabolites are products of enzymes encoded in biosynthetic gene clusters (BGCs), whose prediction from genome sequences has become routine but whose functional roles are rarely tested experimentally. Here, we characterize the BGC repertoire of 225 P. viridiflava isolates from Arabidopsis thaliana and assess BGC contributions to fitness in planta and disease severity. The BGC landscape of P. viridiflava was dominated by non-ribosomal peptide synthetase (NRPS) and NRPS-like BGCs, with one-third of families restricted to a single isolate. Transposon mutagenesis coupled with random barcode transposon sequencing (RB-TnSeq) revealed that the majority of BGCs reduce rather than increase fitness during A. thaliana infection, with the magnitude of the fitness cost varying across host genotypes. This cost could be due to exploitation of public goods by cheater mutant strains. In single-isolate plant infections, where public goods are not available, several BGC families were negatively associated with disease severity, which is positively correlated with bacterial growth in this pathosystem, further indicating that BGCs are generally not beneficial in planta. Our findings reveal extensive and largely uncharacterized biosynthetic potential in populations of P. viridiflava and indicate that candidate metabolites are likely not adaptive for direct interactions with the plant, but perhaps for microbe-microbe interactions either in planta or in other ecological niches. IMPORTANCEBacteria living on plant leaves produce a vast array of chemical compounds, called secondary or specialized metabolites, that can mediate their interaction with the plant host or other microorganisms. Some of these compounds are known to directly influence how bacteria interact with plants, but it has been unclear whether this is a general rule. We studied a large collection of closely related leaf-dwelling bacteria that varied in their ability to cause disease, focusing on leaf-associated Pseudomonas viridiflava--a plant pathogen. We found that very few of the gene clusters responsible for making specialized metabolites improved the ability of the bacteria to colonize Arabidopsis thaliana. On the contrary, carrying these gene clusters often reduced bacterial growth and disease severity in plants. Specialized metabolites may instead primarily be important for interacting with other microbes, different host species, or under environmental conditions we did not test. These are questions that remain for future research.

microbiology↗

bioRxiv2026

A deep-learning-based score to evaluate multiple sequence alignments

Multiple sequence alignment (MSA) inference is a central task in molecular evolution and comparative genomics, and the reliability of downstream analyses, including phylogenetic inference, depends critically on alignment quality. Despite this importance, most widely used MSA methods optimize the sum-of-pairs (SP) score, and relatively little attention has been paid to whether this objective function accurately reflects alignment accuracy. Here, we evaluate the performance of the SP score using simulated and empirical benchmark alignments. For each dataset, we compare alternative MSAs derived from the same unaligned sequences and quantify the relationship between their SP scores and their distances from a reference alignment. We show that the alignment with the optimal SP score often does not correspond to the most accurate alignment. To address this limitation, we develop deep-learning-based scoring functions that integrate a collection of MSA features. We first introduce Model 1, a regression model that predicts the distance of a given MSA from the reference alignment. Across simulated and empirical datasets, this learned score correlates more strongly with true alignment accuracy than the SP score. However, Model 1 is less effective at identifying the best alignment among alternatives. We therefore develop Model 2, which takes as input a set of alternative MSAs generated from the same sequences and predicts their relative ranking. Model 2 more accurately identifies the top-ranking MSA than the SP score, Model 1, and several widely used alignment programs. Using simulations, we show that selecting MSAs based on our approach leads to more accurate phylogenetic reconstructions.

bioinformatics↗

bioRxiv2025

PanGene-O-Meter: Intra-Species Diversity Based on Gene-Content

BackgroundBacterial genome evolution is shaped extensively by horizontal gene transfer, generating gene presence-absence variation that represents a major axis of functional diversity yet is largely invisible to core-genome-based similarity metrics. While genome-wide average nucleotide identity (ANI) remains the most widely used measure of similarity, it considers only the alignable fraction of compared genomes and therefore fails to capture gene-content differences that can have profound phenotypic consequences even among nearly identical strains. ResultsHere we introduce PanGene-O-Meter, a computational framework that quantifies gene-content similarity (GCS) between bacterial genomes by leveraging pan-genome structure and orthology group relationships. Applying PanGene-O-Meter to four large datasets spanning Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, and Helicobacter pylori, we show that GCS captures substantial functional diversity hidden within high-ANI genome pairs, and that gene-content-based trees resolve substructure within sequence types that core-genome analyses cannot. We further introduce GeneContRep, an efficient greedy algorithm for gene-content-based genome dereplication, and validate its utility on 1,000 Klebsiella pneumoniae genomes from two major multidrug-resistant lineages (ST258 and ST307) with matched experimental antimicrobial resistance phenotypes. GeneContRep preserves substantially more clinically relevant diversity than standard ANI-based dereplication, recovering carbapenemase variants, capsule locus types, and resistance determinants that are systematically lost under conventional ANI thresholds. ConclusionsPanGene-O-Meter provides an efficient, complementary, and biologically interpretable framework for quantifying intra-species bacterial diversity, with broad applications in comparative genomics, epidemiological surveillance, and the principled selection of representative genomes for functional studies. PanGene-O-Meter is available at: https://github.com/HaimAshk/PanGene-O-Meter.

microbiology↗

bioRxiv2024

The 1001G+ project: A curated collection of Arabidopsis thaliana long-read genome assemblies to advance plant research

Arabidopsis thaliana was the first plant for which a high-quality genome sequence became available. The publication of the first reference genome sequence almost 25 years ago was already accompanied by genome-wide data on sequence polymorphisms in another accession, or naturally occurring strain. Since then, inventories of genome-wide diversity have been generated at increasingly precise levels. High-density genotype data for A. thaliana, including those from the 1001 Genomes Project, were key to demonstrating the enormous power of GWAS in inbred populations of wild plants, and the comparison of intraspecific polymorphism with interspecific divergence has illuminated many aspects of plant genome evolution. Over the past decade, an increasing number of nearly complete genome sequences have been published for many more accessions. Here, we highlight the diversity of a curated collection of previously published and so far unpublished genome sequences assembled using different types of long reads, including PacBio Continuous Long Reads (CLR), PacBio High Fidelity (HiFi) reads, and Oxford Nanopore Technologies (ONT) reads. This 1001 Genomes Plus (1001G+) resource is being made available at http://1001genomes.org. We invite colleagues with yet unpublished genome assemblies from A. thaliana accessions to contribute to this effort.

genomics↗

bioRxiv2024

Towards an unbiased characterization of genetic polymorphism

Our view of genetic polymorphism is shaped by methods that provide a limited and reference-biased picture. Long-read sequencing technologies, which are starting to provide nearly complete genome sequences for population samples, should solve the problem--except that characterizing and making sense of non-SNP variation is difficult even with perfect sequence data. Here we analyze 27 genomes of Arabidopsis thaliana in an attempt to address these issues, and illustrate what can be learned by analyzing whole-genome polymorphism data in an unbiased manner. Estimated genome sizes range from 135 to 155 Mb, with differences almost entirely due to centromeric and rDNA repeats that are difficult to assemble. The completely assembled chromosome arms comprise roughly 120 Mb in all accessions, but are full of structural variants, largely due to transposable elements. Even with only 27 accessions, a pan-genome coordinate system that includes the resulting variation ends up being [~] 70% larger than the size of any one genome. Our analysis reveals an incompletely annotated mobile-ome: we not only detect several novel TE families, but also find that existing TE annotation is a poor predictor of elements that have recently been active. In contrast to this, the genic portion, or "gene-ome", is highly conserved. By annotating each genome using accession-specific transcriptome data, we find that 13% of all (non-TE) genes are segregating in our 27 accessions, but most of these are transcriptionally silenced. Finally, we show that with short-read data we previously massively underestimated genetic variation of all kinds, including SNPs--mostly in regions where short reads could not be mapped reliably, but also where reads were mapped incorrectly. We demonstrate that SNP-calling errors can be biased by the choice of reference genome, and that RNA-seq and BS-seq results can be strongly affected by mapping reads only to a reference genome rather than to the genome of the assayed individual. In conclusion, while whole-genome polymorphism data pose tremendous analytical challenges, they also have the potential to revolutionize our understanding of genome evolution.

genomics↗

bioRxiv2023

Sensitive clustering of protein sequences at tree-of-life scale using DIAMOND DeepClust

The biosphere genomics era is transforming life science research, but existing methods struggle to efficiently reduce the vast dimensionality of the protein universe. We present DIAMOND DeepClust, an ultra-fast cascaded clustering method optimized to cluster the 19 billion protein sequences currently defining the protein biosphere. As a result, we detect 1.7 billion clusters of which 32% hold more than one sequence. This means that 544 million clusters represent 94% of all known proteins, illustrating that clustering across the tree of life can significantly accelerate comparative studies in the Earth BioGenome era.

genomics↗

bioRxiv2022

Spatially resolved host-bacteria-fungi interactomes via spatial metatranscriptomics

All multicellular organisms are closely associated with microbes, which have a major impact on the health of their host. The interactions of microbes among themselves and with the host take place at the microscale, forming complex networks and spatial patterns that are rarely well understood due to the lack of suitable analytical methods. The importance of high-resolution spatial molecular information has become widely appreciated with the recent advent of spatially resolved transcriptomics. Here, we present Spatial metaTranscriptomics (SmT), a sequencing-based approach that leverages 16S/18S/ITS/poly-d(T) multimodal arrays for simultaneous host transcriptome- and microbiome-wide characterization of tissues at 55-{micro}m resolution. We showcase SmT in outdoor-grown Arabidopsis thaliana leaves as a model system, and found tissue-scale bacterial and fungal hotspots. By network analysis, we study inter- and intra-kingdom spatial interactions among microbes, as well as the host response to microbial hotspots. SmT is a powerful new strategy that will be pivotal to answering fundamental questions on host-microbiome interplay.

genomics↗

bioRxiv2021

Protective host-dependent antagonism among Pseudomonas in the Arabidopsis phyllosphere

The plant microbiome is a rich biotic environment, comprising numerous taxa. The community structure of these colonizers is constrained by multiple factors, including host-microbe and microbe-microbe interactions, as well as the interplay between the two. While much can be learned from pairwise relationships between individual hosts and microbes, or individual microbes with themselves, the ensemble of interrelations between the host and microbial consortia may lead to different outcomes that are not easily predicted from the individual interactions. Their study can thus provide new insights into the complex relationship between plants and microbes. Of particular importance is how strain-specific such plant-microbe-microbe interactions are, and how they eventually affect plant health. Here, we test strain-level interactions in the phyllosphere between groups of co-existing commensal and pathogenic Pseudomonas among each other and with A. thaliana, by employing synthetic communities of genome-barcoded isolates. We found that commensal Pseudomonas prompted a host response leading to a selective inhibition of a specific pathogenic lineage, resulting in plant protection. The extent of plant protection, however, was dependent on plant genotype, indicating that these effects are host-mediated. There were similar genotype-specific effects on the microbe side, as we could pinpoint an individual Pseudomonas isolate as the predominant cause for this differential interaction. Collectively, our work highlights how within-species genetic differences on both the host and microbe side can have profound effects on host-microbe-microbe dynamics. The paradigm that we have established provides a platform for the study of host-dependent microbe-microbe competition and cooperation in the A. thaliana-Pseudomonas system.

plant biology↗

bioRxiv2021

PredictProtein - Predicting Protein Structure and Function for 29 Years

Since 1992 PredictProtein (https://predictprotein.org) is a one-stop online resource for protein sequence analysis with its main site hosted at the Luxembourg Centre for Systems Biomedicine (LCSB) and queried monthly by over 3,000 users in 2020. PredictProtein was the first Internet server for protein predictions. It pioneered combining evolutionary information and machine learning. Given a protein sequence as input, the server outputs multiple sequence alignments, predictions of protein structure in 1D and 2D (secondary structure, solvent accessibility, transmembrane segments, disordered regions, protein flexibility, and disulfide bridges) and predictions of protein function (functional effects of sequence variation or point mutations, Gene Ontology (GO) terms, subcellular localization, and protein-, RNA-, and DNA binding). PredictProteins infrastructure has moved to the LCSB increasing throughput; the use of MMseqs2 sequence search reduced runtime five-fold; user interface elements improved usability, and new prediction methods were added. PredictProtein recently included predictions from deep learning embeddings (GO and secondary structure) and a method for the prediction of proteins and residues binding DNA, RNA, or other proteins. PredictProtein.org aspires to provide reliable predictions to computational and experimental biologists alike. All scripts and methods are freely available for offline execution in high-throughput settings. AvailabilityFreely accessible webserver PredictProtein.org; Source and docker images: github.com/rostlab

bioinformatics↗

bioRxiv2020

A singleton NLR of recent origin causes hybrid necrosis in Arabidopsis thaliana

Hybrid necrosis in plants arises from conflict between divergent alleles of immunity genes contributed by different parents, resulting in autoimmunity. We investigate a severe hybrid necrosis case in Arabidopsis thaliana, where the hybrid does not develop past the cotyledon stage and dies three weeks after sowing. Massive transcriptional changes take place in the hybrid, including the upregulation of most NLR disease resistance genes. This is due to an incompatible interaction between the singleton TIR-NLR gene DANGEROUS MIX 10 (DM10), which was recently relocated from a larger NLR cluster, and an unlinked locus, DANGEROUS MIX 11 (DM11). There are multiple DM10 allelic variants in the global A. thaliana population, several of which have premature stop codons. One of these, which has a truncated LRR domain, corresponds to the DM10 risk allele. The DM10 locus and the adjacent genomic region in the risk allele carriers are highly differentiated from those in the non-risk carriers in the global A. thaliana population, suggesting that this allele became geographically widespread only relatively recently. The DM11 risk allele is much rarer and found only in two accessions from southwestern Spain - a region from which the DM10 risk haplotype is absent - indicating that the ranges of DM10 and DM11 risk alleles may be non-overlapping.

evolutionary biology↗