Biology subjects

Schonhuth, A.

Publications and source records attributed to Schonhuth, A..

9 recordsLinked to original sources

bioRxiv2025

Strain-level metagenomic profiling using pangenome graphs with PanTax

Microbes are omnipresent, thriving in a range of habitats, from oceans to soils, and even within our gastrointestinal tracts. They play a vital role in maintaining ecological equilibrium and promoting the health of their hosts. Consequently, understanding the diversity in terms of strains in microbial communities is crucial, as variations between strains can lead to different phenotypic expressions or diverse biological functions. However, current methods for taxonomic classification from metagenomic sequencing data have several limitations, including their reliance solely on species resolution, support for either short or long reads, or their confinement to a given single species. Most notably, most existing strain-level taxonomic classifiers rely on the sequence representation of multiple linear reference genomes, which fails to capture the sequence correlations among these genomes, potentially introducing ambiguity and biases in metagenomic profiling. Here, we present PanTax, a pangenome graph-based taxonomic profiler that overcomes the short-comings of sequence-based approaches, because pangenome graphs possess the capability to depict the full range of genetic variability present across multiple evolutionarily or environmentally related genomes. PanTax provides a comprehensive solution to taxonomic classification for strain resolution, compatibility with both short and long reads, and compatibility with single or multiple species. Extensive benchmarking results demonstrate that PanTax drastically outperforms state-of-the-art approaches, primarily evidenced by its significantly higher precision or recall at strain level, while maintaining comparable or better performance in other aspects across various datasets. PanTax is an open-source user-friendly tool that is publicly available at https://github.com/LuoGroup2023/PanTax.

bioinformatics↗

bioRxiv2025

PanSpace: Fast and Scalable Indexing for Massive Bacterial Databases

MotivationSpecies identification is a critical task in agriculture, food processing, and health-care. The rapid growth of genomic databases -- driven in part by the increasing investigation of bacterial genomes in clinical microbiology -- has outpaced the capabilities of conventional tools such as BLAST for basic search and query tasks. A key bottleneck in microbiome studies lies in building indexes that allow rapid species identification and classification from assemblies while scaling efficiently to massive resources such as the AllTheBacteria database, thus enabling large-scale analyses to be performed even on a common laptop. ResultsWe introduce PanSpace, the first convolutional neural network-based approach that leverages dense vector (embedding) indexing --- scalable to billions of embeddings --- for indexing and querying massive bacterial genome databases. PanSpace is specifically designed to classify bacterial draft assemblies. Compared to the most recent and competitive tool for this task, PanSpace requires only ~2 GB of disk space to index the AllTheBacteria database, an 8x reduction relative to existing methods. Moreover, it delivers ultra-fast query performance, processing more than 1,000 assemblies in less than two and a half minutes, while preserving the utmost accuracy of state-of-the-art approaches. AvailabilityPanSpace is available at https://github.com/pg-space/panspace.

genomics↗

bioRxiv2024

PanTax: Strain-level taxonomic classification of metagenomic data using pangenome graphs

Microbes are omnipresent, thriving in a range of habitats from oceans to soils and even within our gastrointestinal tracts. They play a vital role in maintaining ecological equilibrium and promoting the health of their hosts. Consequently, understanding the strain diversity within microbial communities is crucial, as variations between strains can lead to distinct phenotypic expressions or diverse biological functions. However, current methods for taxonomic classification from metagenomic sequencing data have several limitations, including their reliance solely on species resolution, support for either short or long reads, or their confinement to a given single species. Most notably, the majority of existing taxonomic classifiers rely solely on a single linear representative genome as a reference, which fails to capture the strain diversity, thereby introducing single-reference biases. Here, we present PanTax, a pangenome graph-based taxonomic classification method that overcomes the shortcomings of single-reference genome-based approaches, because pangenome graphs possess the capability to depict the genetic variability present across multiple evolutionarily or environmentally related genomes. PanTax provides a comprehensive solution to taxonomic classification for strain resolution, compatibility with both short and long reads, and compatibility with single or multiple species. Extensive benchmarking results demonstrate that PanTax drastically outperforms state-of-the-art approaches, primarily evidenced by its significantly higher precision or recall (at both species and strain levels), while maintaining comparable or better performance in other aspects across various datasets. PanTax is a user-friendly open-source tool that is publicly accessible at https://github.com/LuoGroup2023/PanTax.

bioinformatics↗

bioRxiv2023

Hybrid-hybrid correction of errors in long reads with HERO

Although generally superior, hybrid approaches for correcting errors in third-generation sequencing (TGS) reads, using next-generation sequencing (NGS) reads, mistake haplotype-specific variants for errors in polyploid and mixed samples. We suggest HERO, as the first "hybrid-hybrid" approach, to make use of both de Bruijn graphs and overlap graphs for optimal catering to the particular strengths of NGS and TGS reads. Extensive benchmarking experiments demonstrate that HERO improves indel and mismatch error rates by on average 65% (27 95%) and 20% (4 61%). Using HERO prior to genome assembly significantly improves the assemblies in the majority of the relevant categories.

bioinformatics↗

bioRxiv2022

Predicting the prevalence of complex genetic diseases from individual genotype profiles using capsule networks

Diseases that have a complex genetic architecture tend to suffer from considerable amounts of genetic variants that, although playing a role in the disease, have not yet been revealed as such. Two major causes for this phenomenon are genetic variants that do not stack up effects, but interact in complex ways; in addition, as recently suggested, the omnigenic model postulates that variants interact in a holistic manner to establish disease phenotypes. We present DiseaseCapsule, as a capsule network based approach that explicitly addresses to capture the hierarchical structure of the underlying genome data, and has the potential to fully capture the non-linear relationships between variants and disease. DiseaseCapsule is the first such approach to operate in a whole-genome manner when predicting disease occurrence from individual genotype profiles. In experiments, we evaluated DiseaseCapsule on amyotrophic lateral sclerosis (ALS) and Parkinsons disease (PD), with a particular emphasis on ALS because known known to have a complex genetic architecture, so being affected by considerable missing heritability (40%). On ALS, Disease-Capsule achieves 86.9% accuracy on held out test data in predicting disease occurrence, thereby outperforming all other approaches by large margins. Also, DiseaseCapsule required sufficiently less training data for reaching optimal performance. Last but not leaset, the systematic exploitation of the network architecture yielded 922 genes of particular interest, and 644 "non-additive" genes that are crucial factors in DiseaseCapsule, but have no effect within linear schemes.

bioinformatics↗

bioRxiv2022

StrainXpress: strain aware metagenome assembly from short reads

Next-generation sequencing based metagenomics has enabled to identify microorganisms in characteristic habitats without the need for lengthy cultivation. Importantly, clinically relevant phenomena such as resistance to medication, virulence or interactions with the environment can vary already within species. Therefore, a major current challenge is to reconstruct individual genomes from the sequencing reads at the level of strains, and not just the level of species. However, strains of one species can differ only by minor amounts of variants, which makes it difficult to distinguish them. Despite considerable recent progress, related approaches have remained fragmentary so far. Here, we present StrainXpress, as a comprehensive solution to the problem of strain aware metagenome assembly from next-generation sequencing reads. In experiments, StrainXpress reconstructs strain-specific genomes from metagenomes that involve up to more than 1000 strains, and proves to successfully deal with poorly covered strains. The amount of reconstructed strain-specific sequence exceeds that of the current state-of-the-art approaches by on average 26.75% across all data sets (first quartile: 18.51%, median: 26.60%, third quartile: 35.05%).

bioinformatics↗

bioRxiv2022

VeChat: Correcting errors in long reads using variation graphs

Error correction is the canonical first step in long-read sequencing data analysis. The current standard is to make use of a consensus sequence as a template. However, in mixed samples, such as metagenomes or organisms of higher ploidy, consensus induced biases can mask true variants affecting haplotypes of lower frequencies, because they are mistaken as errors. The novelty presented here is to use graph based, instead of sequence based consensus as a template for identifying errors. The advantage is that graph based reference systems also capture variants of lower frequencies, so do not mistakenly mask them as errors. We present VeChat, as a novel approach to implement this idea: VeChat distinguishes errors from haplotype-specific true variants based on variation graphs, which reflect a popular type of data structure for pangenome reference systems. Upon initial construction of an ad-hoc variation graph from the raw input reads, nodes and edges that are due to errors are pruned from that graph by way of an iterative procedure that is based on principles from frequent itemset mining. Upon termination, the graph exclusively contains nodes and edges reflecting true sequential phenomena. Final re-alignments of the raw reads indicate where and how reads need to be corrected. Extensive benchmarking experiments demonstrate that PacBio and ONT reads corrected by VeChat contain 4 to 15, or, respectively, 2 to 10 times less errors than when corrected state of the art approaches. VeChat is implemented in an easy-to-use open-source tool and publicly available at https://github.com/HaploKit/vechat.

bioinformatics↗

bioRxiv2020

MarkerCapsule: Explainable Single Cell Typing using Capsule Networks

Many single cell typing methods require manual annotation which casts problems with respect to resolution of (sub-)types, manpower resources and bias towards existing human knowledge. The integration of heterogeneous data and biologically meaningful interpretation of results are further current key challenges. We introduce MarkerCapsule, which leverages the landmark advantages of capsule networks achieved in their original applications in single cell typing. Thereby, the small amount of labeled data required and the naturally arising, biologically meaningful interpretation of cell types in terms of characteristic gene activity patterns are exemplary strengths, beyond outperforming the state of the art in terms of basic typing accuracy. MarkerCapsule is available at: https://github.com/sumantaray/MarkerCapsule.

bioinformatics↗

bioRxiv2017

Enhancing Sensitivity And Controlling False Discovery Rate In Somatic Indel Discovery Using A Latent Variable Model

Cancer is a genetic disorder in the first place. Therefore, next-generation sequencing (NGS) based discovery of somatically acquired genetic variants has gained widespread attention. Computational prediction of somatic variants, however, is affected by a variety of confounding factors. In addition to the uncertainties that one commonly encounters also in germline variation prediction, such as misplaced and/or inaccurate read alignments, cancer heterogeneity and impure samples significantly add to the issues. Overall, this hampers state-of-the-art indel discovery tools to discover somatic indels at operable performance rates, although they perform excellently when calling germline indels. While affecting all size ranges, both common and cancer-specific problems interfere in particularly unfavorable ways in the prediction of somatic midsize (30-150 bp) insertions and deletions.\n\nHere, we present a latent variable model that can take the major confounding factors and uncertainties into a unifying account. Using this modeling framework, we first demonstrate how to efficiently compute the probability for a (putative) indel to be somatic, thereby resolving a principled computational runtime bottleneck in Bayesian uncertainty quantification. Second, we show how to reliably estimate the allele frequencies for a given list of indels. Third, we also present an intuitive and effective way to control the false discovery rate, an issue in genetic variant discovery that has been found notoriously hard to deal with. As a tool that implements all methodology developed, we present PROSIC (PROcessing Somatic Indel Calls). PROSIC achieves significant improvements in particular in terms of recall when applied to deletion call sheets, as provided by prevalent state-of-the-art tools, in comparison to their integrated somatic indel calling routines.\n\nThe software is publicly available at https://prosic.github.io and can be easily installed via https://bioconda.github.io.

bioinformatics↗