Contact us +49 7071 568 3995

Meet Computomics at ASM Microbe

Computomics at ASM 13-17 June in Atlanta, Georgia

ASM Microbe is the flagship annual meeting of the American Society for Microbiology. It is the largest microbial sciences gathering in the world and will be held in Atlanta, Georgia from 13-17 June, 2024. Join thousands of scientists, researchers and healthcare professionals in your field.

Computomics provides Metagenomics Data Analysis Solutions for bacterial, archaeal, fungal and viral community profiling that enable you to simultaneously identify thousand of species, millions of genes, and diverse biochemical pathways relevant to agriculture, medicine, and the environment. We deliver comprehensive, actionable metagenomics insights through an easy-to-understand interface. From experimental design to data collection, sequencing, analysis, to a written results report - we support you through all stages of your research project.

Josie Bretz and Michelle Hagen will both present a poster. Meet them directly at the poster or reach out to Josie at and Michelle at to schedule your 1:1 meeting.

Poster Sessions:

  • Friday 14 June 2024:
    10:30 AM - 11:30 AM
    4:00 PM - 5:00 PM
  • Saturday 15 June 2024:
    10:30 AM - 11:30 AM
    4:00 PM - 5:00 PM
  • Sunday 16 June 2024:
    10:30 AM - 11:30 AM
    4:00 PM - 5:00 PM

Title: mgPGPT: Metagenomic analysis of plant growth-promoting traits

Metagenomics contributes significantly to plant and soil health estimation. Beneficial microbial strains reduce biotic and abiotic stresses and enhance nutrient acquisition. The recently launched PLaBAse, a comprehensive web resource for plant associated bacteria, addresses the characterisation of their beneficial nature. In this study, we have extended the protein database of the plant growth-promotion traits (PGPT) ontology to facilitate their application for metagenome analysis. The resulting mgPGPT-db, now includes 39,582,183 protein sequences that were computationally curated by incorporating proteins from NCBI, KEGG and AnnoTree. This enhancement integrates the PGPT ontology into our metagenome analysis tools, MEGAN and MORPHEUS. We provide mapping files for the identification of PGPT-related genes through DIAMOND alignment against the new mgPGPT, the NCBI-nr, or the AnnoTree protein databases. The performance comparison of the three metagenomic PGPT assessment approaches on example datasets demonstrates their superiority compared to the original PGPT-db. The mgPGPT-db-based approach achieves the highest read-to-PGPT assignment rates, followed by the AnnoTree-db-based mapping while the NCBInr-db approach returns the highest total aligned and taxonomical assigned read counts. However, the best compromise when considering optimal taxonomic and PGPT annotations provides the AnnoTree-based approach with a still acceptable runtime.Finally, we have compared the inferred PGPT content of several samples taken from different environments and show plant specific PGPT clustering.

Josie Bretz, Bioinformatics Analyst


Title: Deciphering the Soil Microbiome's Marker Taxa under Drought Stress Using Interpretable Machine Learning
In the face of climate-induced droughts impacting crop yields and food security novel strategies are needed to enhance crop productivity and resistance. Metagenomics gains acknowledgement in terms of identification of stress-tolerant and -reducing microbial communities that could be applied for microbiome transplantation, bioinoculation, and for early-state drought stress detection in the fields. By diving into the world of interpretable Machine Learning and employing Shapley Additive Explanation (SHAP) values, alongside conventional Differential Abundance Analysis (DAA) methods, we aim to identify marker taxa and improve the prediction of drought stress in the soil microbiome of varying crop species. Our results show that both Machine Learning-based SHAP values and the metagenomics approach of DAA methods provide similar information. They exhibit their potential as complementary approaches for identifying marker taxa, as illustrated in Fig. 1. But, a Random Forest Classifier, trained on a diverse set of soil metagenome samples of different grass species, can achieve high accuracy in the prediction of drought stress of about 92.3 % at the genus level, as displayed in Tab. 1. Our drought stress classifier demonstrates its generalization capabilities across various drought stress levels and grass lineages. We emphasize the potential to detect drought stress in the soil metagenome using an optimized and generalized location-based Machine Learning classifier. Through the identification of drought-specific marker taxa, this approach holds promising implications for microbe-assisted plant breeding and agricultural management programs. It contributes towards the development of sustainable agricultural practices, essential for securing global food supplies in the face of climate change.

Michelle Hagen, Bioinformatics Analyst

Find the full program here.

We’re looking forward to meeting you in Atlanta!

Share on

Get in touch with us