• Susannah Tringe, Susannah Tringe, Lawrence Berkeley National Laboratory

Plants roots and the soil they grow in are heavily colonized with microbes that play critical roles in nutrient cycling and transport as well as influencing plant growth and health. Molecular methods including DNA sequencing have begun to elucidate the forces governing the assembly and maintenance of plant and soil microbial communities, offering the opportunity for these microbial communities to be nurtured and manipulated to promote plant growth and health as well as soil health and ecosystem functions.
We have combined omics methods, biogeochemical assays, and gas flux measurements to investigate the factors influencing greenhouse gas emissions from natural and managed wetland systems. By integrating these datasets we find that gas fluxes represent a complex interplay of biological, chemical, and physical factors that vary across habitats. Our results suggest considerable heterogeneity in fluxes even in physically proximate locations that have implications for the success of wetland preservation and restoration as a carbon storage strategy, particularly in the context of sea level rise.
In agricultural systems, we find that different plant compartments (e.g. rhizosphere and root endosphere) harbor unique and dynamic microbial communities heavily influenced by the soil, surrounding environment and host genotype. Abiotic stress, such as drought and low nitrogen, can alter both the composition of these communities and their interactions with each other and the plant. Our sequence-based characterizations of plant-associated communities, leveraging a variety of bioinformatic tools, have identified key populations that structure the community and respond dynamically to environmental changes, representing potential targets for improvement of plant resilience.

• Célio Dias Santos-Júnior, Célio Dias Santos-Júnior, Fudan University
• Marcelo Torres, Marcelo Torres, University of Pennsylvania
• Yiqian Duan, Yiqian Duan, Fudan University
• Cesar de la Fuente Nunez, Cesar de la Fuente Nunez, University of Pennsylvania
• Luis Pedro Coelho, Luis Pedro Coelho, Queensland University of Technology

Small proteins, crucial across all life domains, have been overlooked in large-scale microbiome studies due to limitations in both wet lab and bioinformatics techniques. In particular, it is difficult to predict them without generating numerous false positives, and functional predictions based on homology fail without closely-related homologs. Recently, studies have begun addressing these challenges, with improved methods for managing small protein data in metagenomic analyses.

We tackled this by analyzing sequences shared across multiple metagenomes, to increase confidence in predictions. This method was applied in creating the Global Microbial smORF Catalogue, which includes almost one billion sequences. This is accessible online for users to identify homologs to smORFs identified in their own studies.

Additionally, we used machine learning to filter out false positives effectively, particularly in identifying active sequences within specific functional classes like antimicrobial peptides (AMPs). For this task, we designed macrel to optimize for high precision, albeit at the potential cost of lower recall. Macrel was used to generate a catalog of one million potential AMPs from extensive genomic and metagenomic data, a dataset we termed AMPsphere.

Experimental validation of these methods included synthesizing and testing 100 AMPs. In total, 79 showed activity against pathogens or commensals. Some peptides also demonstrated efficacy comparable to the clinical antimicrobial, polymyxin B, in a preclinical mouse model, underscoring the potential of these novel bioinformatic approaches to contribute significantly to discovering novel antibiotics.

• James Morton, James Morton, Gutz Analytics

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by heterogeneous cognitive, behavioral and communication impairments. Disruption of the gut–brain axis (GBA) has been implicated in ASD although with limited reproducibility across studies. In this study, we developed a Bayesian differential ranking algorithm to identify ASD-associated molecular and taxa profiles across 10 cross-sectional microbiome datasets and 15 other datasets, including dietary patterns, metabolomics, cytokine profiles and human brain gene expression profiles. We found a functional architecture along the GBA that correlates with heterogeneity of ASD phenotypes, and it is characterized by ASD-associated amino acid, carbohydrate and lipid profiles predominantly encoded by microbial species in the genera Prevotella, Bifidobacterium, Desulfovibrio and Bacteroides and correlates with brain gene expression changes, restrictive dietary patterns and pro-inflammatory cytokine profiles. The functional architecture revealed in age-matched and sex-matched cohorts is not present in sibling-matched cohorts. We also show a strong association between temporal changes in microbiome composition and ASD phenotypes. In summary, we propose a framework to leverage multi-omic datasets from well-defined cohorts and investigate how the GBA influences ASD.

• Xiaofang Jiang, Xiaofang Jiang, NLM/NIH
• Keith Dufault-Thompson, Keith Dufault-Thompson, National Library of Medicine
• Brantley Hall, Brantley Hall, Department of Cell Biology and Molecular Genetics

The gut microbiome plays crucial roles in animal health and metabolism, including the biotransformation of host and diet-derived metabolites. The microbial reduction of bilirubin, a heme degradation product, to urobilinogen is a key process in maintaining bilirubin homeostasis in animals. This study employs phylogenetic and metagenomic mining approaches to unveil a novel family of gut-adapted bilirubin reductase enzymes within the Old Yellow Enzyme (OYE) family.

Through an integrated analysis combining experimental screening, comparative genomics, and advanced computational methodologies, we identified and characterized a putative bilirubin reductase enzyme family in anaerobic microbes associated with the gut. Through structural modeling, ancestral sequence reconstruction, and targeted mutation experiments, we confirmed the specificity and function of these enzymes, delineating them from other members of the OYE family. Our findings reveal three distinct forms of bilirubin reductase, characterized by unique domain compositions, that form separate clades within the enzyme's phylogeny.

Our analysis of 1373 gut metagenomes across 132 animal species illuminated the evolutionary divergence and niche-specific associations of the bilirubin reductase clades. We found that bilirubin reductase was significantly enriched in the anaerobic niche of the lower gut in multiple animals, being nearly absent in their upper gastrointestinal tracts. The broader distribution of bilirubin reductase clades highlights clear patterns of co-evolution with their animal hosts, underscoring the ecological and evolutionary interplay between gut microbes and their vertebrate hosts.

• Mohammadali Serajian, Mohammadali Serajian, Unievrsity of Florida
• Simone Marini, Simone Marini, University of Florida
• Jarno N. Alanko, Jarno N. Alanko, University of Helsinki
• Noelle R. Noyes, Noelle R. Noyes, University of Minnesota
• Mattia Prosperi, Mattia Prosperi, University of Florida
• Christina Boucher, Christina Boucher, University of Florida

We develop a robust machine learning classifier using both linear and nonlinear models (i.e., LASSO logistic regression (LR) and random forests (RF)) to predict the phenotypic resistance of \emph{Mycobacterium tuberculosis} (MTB) for a broad range of antibiotic drugs. We use data from the CRyPTIC consortium to train our classifier, which consists of whole genome sequencing and antibiotic susceptibility testing (AST) phenotypic data for 13 different antibiotics. To train our model, we assemble the sequence data into genomic contigs, identify all unique 31-mers in the set of contigs, and build a feature matrix M, where M[i,j] is equal to the number of times the i-th 31-mer occurs in the j-th genome. Due to the size of this feature matrix (over 350 million unique 31-mers), we build and use a sparse matrix representation. Our method, which we refer to as MTB++, leverages compact data structures and iterative methods to allow for the screening of all the 31-mers in the development of both LASSO LR and RF. MTB++ is able to achieve high discrimination (F-1 greater than 80%) for the first-line antibiotics. Moreover, MTB++ had the highest F-1 score in all but three classes and was the most comprehensive since it had an F-1 score greater than 75% in all but four (rare) antibiotic drugs. We use our feature selection to contextualize the 31-mers that are used for the prediction of phenotypic resistance, leading to some insights about sequence similarity to genes in MEGARes.

• Ba Thong Nguyen, Ba Thong Nguyen, University of Hawaii at Manoa
• Shaoqiu Chen, Shaoqiu Chen, University of Hawaii at Manoa
• Donna Lee Kuehu, Donna Lee Kuehu, University of Hawaii at Manoa
• Isam Ibrahim, Isam Ibrahim, University of Hawaii at Manoa
• Yujia Qin, Yujia Qin, University of Hawaii at Manoa
• Youping Deng, Youping Deng, University of Hawaii at Manoa

Background: Identifying factors underlying resistance to immune checkpoint therapy (ICT) is still challenging. Intratumor microbes (bacteria, fungi, and viruses) are found in multiple tumor tissues of many cancers. In this study, we examined the intratumor microbes of lung patients under immune checkpoint inhibitors.
Methods: We downloaded the whole exome sequencing (WXS) that contains primary tumor and non-tumor non-small cell lung cancer (NSCLC) with clinical data downloaded from Genomes and Phenotypes (dbGaP) databases. The data was collected from three NSCLC with ICT response cohorts (phs002244, phs000980 and phs001940), including 44 response (R) samples and 51 nonresponse (NR) samples. The microbes’ abundance, diversity and significant microbes were extracted through machine learning and microbiome analysis.
Results: The whole exome sequencing (WXS) of 95 patient’s data (NR, 51; R,44) were obtained and analyzed from Feb 2023 to April 2024. After cleaning up, and microbiome analysis data, we found significant significantly higher alpha diversity in response group compare with non-response group in three type microbes in tumor samples (p<0.05) while not significant found in non-tumor tissues. Through different microbes’ analysis, we found bacterium (Lactobacillus gasseri), fungi (Aspergillus_versicolor, GS01_phy_Incertae_sedis_sp) and viruses (Alphabaculovirus, and Mardivirus) are top significant species and genus in response group compared to non-response group in tumor tissues. We found top abundance species and genus of bacteria (Lactobacillus gasseri, and Ralstonia solanacearum), fungi (Aspergillus, Fungi_gen_Incertae_sedis) and viruses (Alphabaculovirus and Betapartitivirus) in tumor samples.
Conclusion: Together, these microbes data provide important implications for the treatment of lung cancer with immune checkpoint inhibitors.

• Jithin Sunny, Jithin Sunny, Queen's University
• Sabhjeet Kaur, Sabhjeet Kaur, Queen's University
• Jeremie Alexander, Jeremie Alexander, Queen's University
• George C. Dicenzo, George C. Dicenzo, Queen's University

This study utilizes comprehensive bioinformatics approaches to investigate the gut bacterial population of mealworms and superworms and to mine for enzymes potentially involved in plastic degradation. A total of 46 metagenomes were assembled, annotated, and analyzed to characterize the bacterial population and identify taxa differentially abundant between insects fed plastics and those not fed plastics. Alpha and beta diversity metrics were first used to examine global differences between the diet groups followed by non-symmetric analysis to explain the variation in data. Binning of the metagenome assemblies led to the generation of 153 metagenome-assembled genomes (MAGs). Metabolic pathway analysis was performed for these MAGs to observe the gene counts involved in aromatic compound degradation genes belonging to the butanoate, and propanoate metabolic pathways amongst others. To further explore genes potentially associated with plastic biodegradation, we annotated all the metagenomes and extracted a non-redundant set of ~105,000 proteins. The non-redundant set of exported proteins included 129 putative laccases, which were of interest as previous studies have implicated this protein family in plastic degradation. We therefore performed sequence and structural analyses to explore the properties of the putative laccases identified in our study. Features were computed using site and domain-based information along with residue and enzyme backbone based structural similarity. Three different clustering methods along with evaluation metrics were employed to evaluate enzymes showing high similarity to laccases previously suggested to be active on plastics. Overall, this research employs different bioinformatics techniques to understand the bacterial groups and enzymes involved in plastic degradation.