Biology subjects

Vandamme, P.

Publications and source records attributed to Vandamme, P..

3 recordsLinked to original sources

bioRxiv2024

Pre-trained Maldi Transformers improve MALDI-TOF MS-based prediction

For the last decade, matrix-assisted laser desportion/ionization time-of-flight mass spectrometry (MALDI-TOF MS) has been the reference method for species identification in clinical microbiology. Hampered by a historical lack of open data, machine learning research towards models specifically adapted to MALDI-TOF MS remains in its infancy. Given the growing complexity of available datasets (such as large-scale antimicrobial resistance prediction), a need for models that (1) are specifically designed for MALDI-TOF MS data, and (2) have high representational capacity, presents itself. Here, we introduce Maldi Transformer, an adaptation of the state-of-the-art transformer architecture to the MALDI-TOF mass spectral domain. We propose the first self-supervised pre-training technique specifically designed for mass spectra. The technique is based on shuffling peaks across spectra, and pre-training the transformer as a peak discriminator. Extensive benchmarks confirm the efficacy of this novel design. The final result is a model exhibiting state-of-the-art (or competitive) performance on downstream prediction tasks. In addition, we show that Maldi Transformers identification of noisy spectra may be leveraged towards higher predictive performance. All code supporting this study is distributed on PyPI and is packaged under: https://github.com/gdewael/maldi-nn

bioinformatics↗

bioRxiv2022

The GEN-ERA toolbox: unified and reproducible workflows for research in microbial genomics .

BackgroundMicrobial culture collections play a key role in taxonomy by studying the diversity of their accessions and providing well characterized strains to the scientific community for fundamental and applied research. These microbial resource centers thus need to implement new standards in species delineation, including whole-genome sequencing and phylogenomics. In this context, the genomic needs of the Belgian Coordinated Collections of Microorganisms (BCCM) were studied, resulting in the GEN-ERA toolbox. The latter is a unified cluster of bioinformatic workflows dedicated to both bacteria and small eukaryotes (i.e. yeasts). FindingsThis public toolbox allows researchers without a specific training in bioinformatics to perform robust phylogenetic analyses. Hence, it facilitates all steps from genome downloading and quality assessment, including genomic contamination estimation, to tree reconstruction. It also offers workflows for average nucleotide identity comparisons and metabolic modeling. Technical detailsNextflow workflows are launched by a single command and are available on the GEN-ERA GitHub repository (https://github.com/Lcornet/GENERA). All the workflows are based on Singularity containers to increase reproducibility. TestingThe toolbox was developed for a diversity of microorganisms, including bacteria and fungi. It was further tested on an empirical dataset of 18 (meta)genomes of early-branching Cyanobacteria, providing the most up-to-date phylogenomic analysis of the Gloeobacterales order, the first group to diverge in the evolutionary tree of Cyanobacteria. ConclusionThe GEN-ERA toolbox can be used to infer completely reproducible comparative genomic and metabolic analyses on prokaryotes and small eukaryotes. Although designed for routine bioinformatics of culture collections, it can also be useful for other applications, as shown by our case study on Gloeobacterales.

bioinformatics↗

bioRxiv2022

Cyclitol secondary metabolism is a central feature of Burkholderia leaf symbionts

The symbioses between plants of the Rubiaceae and Primulaceae families with Burkholderia bacteria represent unique and intimate plant-bacterial relationships. Many of these interactions have been identified through PCR-dependent typing methods, but there is little information available about their functional and ecological roles. We assembled seventeen new endophyte genomes representing endophytes from thirteen plant species, including those of two previously unknown associations. Genomes of leaf endophytes belonging to Burkholderia s.l. show extensive signs of genome reduction, albeit to varying degrees. Except for one endophyte, none of the bacterial symbionts could be isolated on standard microbiological media. Despite their taxonomic diversity, all endophyte genomes contained gene clusters linked to the production of specialized metabolites, including genes linked to cyclitol sugar analog metabolism and in one instance non-ribosomal peptide synthesis. These genes and gene clusters are unique within Burkholderia s.l. and are likely horizontally acquired. We propose that the acquisition of secondary metabolite gene clusters through horizontal gene transfer is a prerequisite for the evolution of a stable association between these endophytes and their hosts.

microbiology↗