Posters - Schedules
Poster presentations at ISMB/ECCB 2021 will be presented virtually. Authors will pre-record their poster talk (5-7
minutes) and will upload it to the virtual conference platform site along with a PDF of their poster beginning July 19
and no later than July 23. All registered conference participants will have access to the poster and presentation
through the conference and content until October 31, 2021. There are Q&A opportunities through a chat
function and poster presenters can schedule small group discussions with up to 15 delegates during the conference.
Information on preparing your poster and poster talk are available at:
https://www.iscb.org/ismbeccb2021-general/presenterinfo#posters
Ideally authors should be available for interactive chat during the times noted below:
View Posters By Category
Session A: Sunday, July 25 between 15:20 - 16:20 UTC |
Session B: Monday, July 26 between 15:20 - 16:20 UTC |
|---|---|
Session C: Tuesday, July 27 between 15:20 - 16:20 UTC |
Session D: Wednesday, July 28 between 15:20 - 16:20 UTC |
|---|---|
Session E: Thursday, July 29 between 15:20 - 16:20 UTC |
|---|
Short Abstract: Root-associated microbiota, such as rhizospheric and endophytic bacterial communities, have a great effect on plant growth. Some plant growth-promoting bacteria have been applied to the field as biofertilizers, but the effect to improve crop production is unstable. For application to agriculture, it is important to understand complex biological interaction surrounding the roots and inoculation effect on the endophytic microbial communities.
We collected a set of growth and endophytic microbiome data of individual tomato (Solanum lycopersicum) plants grown under high temperature and various synthetic microbial community (SynCom) conditions. We used the resulting data to predict plant phenotype or endophytic taxonomic profile by SynCom profile (combination of inoculants) and environmental conditions using linear regression models.
From the taxonomic profile prediction model, the frequency of several bacterial genera could be predicted from the combination of inoculants and environmental conditions, and the important factors affecting the frequency change were identified. From the growth prediction model, several inoculants were found as candidate biomarkers to improve tomato growth.
Short Abstract: The microbiome has been implicated as a potential driver in a variety of cancers, including breast cancer [Hieken 2016]. There are numerous studies, with associated microbial sequencing data, available on this topic. [Urbaniak 2014; Xuan 2014; Hieken 2016]. However, microbiome analysis is a rapidly developing field and older analytical methods rapidly become deprecated. In the work presented here we have performed a meta-analysis of the raw data from available studies, with the intended goal of assessing (1) the reproducibility of the original findings and (2) assessing breast cancer-specific microbial signatures that occur across study cohorts [Urbaniak 2014; Xuan 2014; Hieken 2016; Chan 2016]. Using the DADA2 pipeline, an amplicon sequence variant approach providing high taxonomic resolution, specificity, and lower false positives than the mixture of now-outdated bioinformatic approaches used in the original studies [Prodan 2020], we have identified major improvements in analysis approaches as well as novel findings that were not revealed in the original reports. We anticipate that these new research results arising from a thorough meta-analysis of existing datasets using modern best-practices and up-to-date databases will provide researchers vital information surrounding the relationship between breast cancer and its associated microbial communities.
Short Abstract: The honey bee (Apis mellifera) gut microbiome has been characterized using 16S rRNA amplicon and shotgun metagenomic sequencing. It mainly consists of a conserved microbial community that directly impacts bee health and immunity. Our study aimed to identify the most informative sample and sequencing approach to characterize the honey bee gut microbiome. Whole bee, abdomen, gut and hindgut from individual bees, or from a pool of ten bees, were analyzed using 16S rRNA amplicon and metagenomic shallow and deep sequencing. At the genus level, samples from the individual or pooled bees resulted in similar findings for all sequencing approaches. Kraken 2 analyses accurately identified members of the honey bee microbiome. However, metagenomic sequencing was found to be advantageous at the species level. Species rarefaction analysis revealed that deep-sequenced samples discovered species at a greater rate than shallow-sequenced samples, which indicated that taxa abundance information is gained through increased sequencing depth. While data obtained from whole bee and abdomen were informative, the gut and hindgut yielded a higher proportion of microbial reads compared to host reads. Therefore, sequencing gut or hindgut samples at a greater depth is an ideal approach for bee microbiome analyses to address questions related to bee health.
Short Abstract: The human gut microbiome is a highly diverse ecosystem which harbors thousands of microbial species. Changes in its composition and function have been linked to conditions such as inflammatory bowel disease, type-1 diabetes, and many other diseases. However, the exact mechanisms of how the microbiota influences health is still elusive, due to our limited understanding of the functional potential encoded in complex microbial genomes. This study aims to characterize the role of the human gut microbiome in type-1 diabetes. We used DIABIMMUNE infant gut microbiome data as a case study. We developed a custom metagenomics annotation pipeline centered around DeepFRI, a deep learning method that combines sequence-based LSTM and 3D structure-based Graph Convolutional Networks. We generated a sequence catalog comprising 2,255 genomes and 1.9M non-redundant microbial genes. With DeepFRI predictions, there was an increase in the annotation coverage weighted by relative abundance. In addition, we carried out pan-genome analysis of 42 species. This analysis revealed a diversity of functional enrichment between the core and accessory genomes. We provide a computational methodology for annotation of human gut metagenome sequences with high coverage. This work contributes to the understanding of the functional signature of the human gut microbiome in health and disease.
Short Abstract: The microflora of the plant rhizosphere (the soil environment in the vicinity of plant roots) has a vital role in plant health and growth. Thus an accurate understanding of the rhizosphere microbiota is of great importance for plant physiology and agronomy. Generative models, such as variational autoencoders (VAEs) and generative adversarial networks (GANs), can imitate the phenomena of interest and provide helpful information to understand the data generation mechanisms. The recent accumulation of the 16S amplicon sequencing-based microbiome data brings the opportunity to apply this generative modeling approach to microbial ecology. This study attempted to build generative models for microbial community profiles of both the plant rhizosphere and soil environments using a comprehensive 16S amplicon sequencing dataset retrieved from the publicly available repositories. Learning models with regularization terms for microbial ecological relationships (characteristics of microbial co-occurrence networks) successfully reduced the generalization error. We could also estimate some environment-specific microbial community profiles based on the constructed models. In this conference, we would like to discuss the possibility of latent space interpretation of microbiome generative models and the applicability of these techniques to the performance improvement of classification tasks using microbial abundance as explanatory variables.
Short Abstract: Fibrillar adhesins are a novel defined class of bacterial adhesins, which play an important role in host-pathogen interactions and are found in a wide range of bacterial species. The name ‘fibrillar adhesins’ originates from their filamentous shape, which is based on repeating protein domains forming a rod-like stalk. An additional adhesive domain enables the binding to protein ligands, carbohydrates or even ice crystals.
By analysing fibrillar adhesins in depth, we identified several characteristics such as their amino acid composition or cell surface anchoring mechanisms. Knowing these characteristics, we developed a RandomForest based discovery approach to detect fibrillar adhesins. By applying this approach to the Firmicutes and Actinobacteria UniProt reference proteomes, many novel fibrillar adhesins were predicted, including proteins without known adhesive domains. These are very interesting candidates that may contain novel adhesive domains awaiting discovery. Investigating these candidates by predicting the domain structure and searching for homologous sequences and similar structures in the PDB database, led to the discovery of potential families of novel adhesive domains. Three of these potentially novel adhesive domains are found in proteins of pathogenic Bacillus, Enterococcus or Listeria species.
Short Abstract: Summary: MMseqs2 taxonomy is a new tool to assign taxonomic labels to metagenomic contigs. It extracts all pos- sible protein fragments from each contig, quickly retains those that can contribute to taxonomic annotation, assigns them with robust labels and determines the contig’s taxonomic identity by weighted voting. Its fragment extraction step is suitable for the analysis of all domains of life. MMseqs2 taxonomy is 2-18x faster than state-of-the-art tools and also contains new modules for creating and manipulating taxonomic reference databases as well as reporting and visualizing taxonomic assignments.
Availability and implementation: MMseqs2 taxonomy is part of the MMseqs2 free open-source software package available for Linux, macOS and Windows at mmseqs.com.
Short Abstract: Fungi are vital elements in several critical ecological functions, ranging from organic matter decomposition to symbiotic associations with plants. Moreover, fungi naturally inhabit the human microbiome and can be causative agents of human infections. Although many machine learning methods have been proposed for the taxonomic classification of fungal ITS sequences, to the best of our knowledge, none of them fully explore the information provided in the taxonomic tree hierarchy when building their models. This in turn, leads to lower generalization power and higher classification errors. Here, we introduce HiTaC, a robust, hierarchical machine learning model for accurate ITS classification, which requires a small amount of data for training and handles imbalanced datasets. HiTaC was thoroughly evaluated with the established TAXXI benchmark, characterizing its ability to correctly classify fungal ITS sequences of varying lengths and a range of identity differences between the training and test data. HiTaC outperforms state-of-the-art methods, consistently achieving higher accuracy and sensitivity across different taxonomic ranks, improving accuracy by 5.2% and sensitivity by 6.9% over top methods in the most noisy dataset available on TAXXI. A user-friendly QIIME2 plugin allows users to include HiTaC in existing pipelines easily. HiTaC is publicly available and freely accessible at gitlab.com/dacs-hpi/hitac.
Short Abstract: An increasing number of raw sequencing datasets of the bee microbiome have become publicly accessible on the Sequence Read Archive (SRA). There are currently over nine thousand datasets in SRA from over 140 Apoidea hosts, which originate from 238 different geographical locations. This massive amount of data has brought an increasing complexity to retrieve relevant datasets and the ability to stay up to date. The goal of the Bee Microbiome consortium (beebiome.org/beta/consortium/members) is to facilitate access to metadata of publicly available datasets for bee microbiome scientists, and promote outreach with stakeholders and the general public. Our team’s goal is to advance the development of a bee microbiome community portal and improve its functionality with the Tripal web framework which allows for FAIR data sharing and collaboration, as well as portability with Docker. Tripal has been used in over 30 databases and has been in development for over 11 years, and is supported by a large and active community. We are also exploring integration with other analysis tools and Galaxy workflows via the Tripal extension modules, for example with BioNet Alberta Galaxy or public Galaxies that will enable users to streamline the analysis of the retrieved data.
Short Abstract: Metagenomic binning algorithms computationally group scaffolds, assembled from whole-genome shotgun (WGS) reads, according to the species of origin. Current tools produce many low-completeness bins, especially on datasets with few samples. Here, we built machine learning models that integrate long-range DNA interaction information provided by Hi-C sequencing to improve binning. The resulting software, Metabat-HiC, automatically recruits previously unbinned scaffolds into the correct bins and merges bins that belong to the same genome. Applying the software to a real-world cat microbiome dataset led to an average 6% increase in genome completeness among all bins without significantly increased contamination levels. We are benchmarking Metabat-HiC on more datasets as well as against other popular binning algorithms such as ProxiMeta and bin3C.
Short Abstract: We have developed MUFFIN, a complete metagenomic workflow producing high-quality metagenome-assembled genomes (MAGs) and their annotations using short (Illumina) and long (nanopore) reads.
The core of metagenomics studies is to obtain high-quality and high completeness genomes assemblies. Those assemblies serve as the foundation for the functional analysis, classification, and detection of interactions between the organisms of a microbiome.
MUFFIN produces high-quality and high-completeness bins using hybrid assembly, differential coverage binning methods, and bins refinement.
The workflow also produces the taxonomic classification (GTDB) and annotation (EggNOG). It provides summary files for the classification, bin quality, and interactive web pages for the KEGG pathways obtained during the annotation. Finally, MUFFIN produces functional pathway predictions, and if RNA-seq data is provided, de novo metatranscript annotations across the metagenomic sample and for each bin.
The workflow is written using Nextflow, a workflow orchestration software, to achieve high reproducibility, platform-independent, fast, and straightforward analysis. MUFFIN is easy to install and will help researchers obtain great genome assemblies from metagenome sequencing data with high reproducibility.
MUFFIN is available on GitHub under GNUv3 license: github.com/RVanDamme/MUFFIN, the test data is available on the OSF: osf.io/m5czv/ and the paper is available in PLOS: doi.org/10.1371/journal.pcbi.1008716
Short Abstract: Whole microbiome RNASeq (metatranscriptomics) is a powerful technology to functionally interrogate microbial communities. A key challenge is how best to process and interpret these complex datasets. In typical applications, single metatranscriptomic datasets comprise of millions of sequence reads. These reads are filtered for low quality and contaminants, before being annotated with taxonomic and functional labels, and collated to generate global bacterial gene expression profiles. We present MetaPro, a flexible, scalable metatranscriptomic data analysis pipeline in a Docker framework. MetaPro starts with raw sequence read inputs (single or paired end reads) and processes them through a tiered series of filtering, assembly, and annotation steps. In addition to yielding a final list of bacterial genes and their relative expression, MetaPro delivers a taxonomic breakdown based on the consensus of complementary prediction algorithms, together with a focused breakdown of enzymes, readily visualized through Cytoscape. We benchmark the performance of MetaPro against two current pipelines and demonstrate improved performance and functionality. MetaPro represents an effective integrated solution for processing and analysing metatranscriptomic datasets. MetaPro, together with an established tutorial that has been developed for educational purposes is made freely available at github.com/ParkinsonLab/MetaPro. The software is freely available under the GNU general public license v3.
Short Abstract: DNA de novo assembly, a method that can reconstruct longer stretches of DNA (contigs) from short sequencing reads, is increasingly used for ancient DNA (aDNA) metagenomics data.
However, while the source of a modern sample is often unambiguous, ancient samples from archeological origins are often formed from a mixture of ancient DNA and modern contaminants. It is therefore essential to be able to distinguish between truly ancient contigs and contigs assembled from contaminant DNA, possibly originating from excavation, storage environment, or any other form of modern interference. aDNA main characteristic damage pattern, deamination, is usually one of the key elements advocating for the authenticity of an aDNA sequence. Yet tools designed for inspecting and filtering aDNA damage do not scale well with the sheer amount of contigs generated by metagenomic de novo assembly. To address these challenges and limitations, we designed PyDamage, an automated approach for damaged contig identification and aDNA damage estimation.
PyDamage uses a statistical approach to discriminate between truly ancient contigs, and contigs originating from modern contamination.
We tested PyDamage, both on simulated ancient DNA sequencing data, and real archeological samples, and demonstrated its ability to automatically retrieve and identify contigs bearing deamination based DNA damage.
Short Abstract: Background: Metagenomic sequencing has led to the identification and assembly of many new bacterial genome sequences. These bacteria often contain plasmids: usually small, circular double-stranded DNA molecules that may transfer across bacterial species and confer antibiotic resistance and are less studied and understood. Part of the reason for this is the lack of computational tools enabling the analysis of plasmids in metagenomic samples.
Results: In order to assist in the study of plasmids we developed SCAPP (Sequence Contents-Aware Plasmid Peeler) - an algorithm and tool to assemble plasmid sequences from metagenomic sequencing. We compared the performance of SCAPP to Recycler and metaplasmidSPAdes on simulated metagenomes, real human gut microbiome samples, and a human gut plasmidome dataset that we generated. We also created plasmidome and metagenome data from the same cow rumen sample and used the parallel sequencing data to create a novel assessment procedure. In almost all cases SCAPP outperformed Recycler and metaplasmidSPAdes across this wide range of datasets.
Availability: SCAPP is an easy to use Python package that enables the assembly of full plasmid sequences from metagenomic samples. SCAPP is open-source software available from: github.com/Shamir-Lab/SCAPP.
This study is in press in BMC Microbiome.
Short Abstract: Summary: SpacePHARER (CRISPR Spacer Phage-Host Pair Finder) is a sensitive and fast tool for de novo prediction of phage–host relationships via identifying phage genomes that match CRISPR spacers in genomic or metagenomic data. SpacePHARER gains sensitivity by comparing spacers and phages at the protein level, optimizing its scores for matching very short sequences, and combining evidence from multiple matches, while controlling for false positives. We demonstrate that SpacePHARER has 1.4 to 4 times higher sensitivity than the state-of-the-art BLASTN-based method, while being around 47 times faster, by searching a comprehensive spacer list against all available complete phage genomes on NCBI. We hope that SpacePHARER will help uncovering many unknown phage-host relationships and facilitate further understanding of phage-host co-evolution.
Availability and implementation: SpacePHARER is available as an open-source (GPLv3), user-friendly command-line software for Linux and macOS: github.com/soedinglab/spacepharer.
Short Abstract: Taxonomic metagenome profiling aims to predict the identity and relative abundances of taxa in a given whole genome sequencing metagenomic sample. A recent surge in computational methods that aim to accomplish this, called taxonomic profilers, has motivated community-driven efforts to create standardized benchmarking datasets, standardized taxonomic profile formats, and benchmarking platforms to assess tool performance. Here we report the development of Tampa (Taxonomic metagenome profiling evaluation), a robust and easy-to-use method that allows scientists to easily interpret and interact with taxonomic profiles produced by the many different taxonomic profiler methods. We demonstrate Tampa's ability by illuminating the critical biological differences between samples and conditions otherwise missed by commonly utilized metrics. We plan to apply Tampa to CAMI data1. Additionally, we show that Tampa can enable biologists to effectively choose the most appropriate profiling method to use on their real data. When ground truth taxonomic profiles are available, we show how Tampa can augment existing benchmarking platforms such as OPAL. Tampa will be provided in a platform-independent fashion via Bioconda and integrated into the Galaxy Toolshed. Tampa will allow scientists to quickly contextualize, assess, and extract insight from taxonomic profiles instead of relying primarily on statistical summaries or manual manipulation.
Short Abstract: "Inflammatory bowel diseases (IBD) constitute a spectrum of chronic inflammatory disorders that recurrently affect the gastrointestinal tract. Ulcerative colitis (UC) and Crohn’s disease (CD) are the two main clinically defined manifestations of IBD. A growing number of reports showing the alteration of gut microbiota in subjects with IBD indicate the potential benefit of exploiting metagenomics for non-invasive IBD diagnostics.
In order to investigate the diagnostic potential of metagenomics data in regard to IBD, we conducted the crowdsourced sbv IMPROVER Metagenomics Diagnosis for Inflammatory Bowel Disease Challenge, open to the worldwide scientific community, from September 2019 until March 2020.
Participants were provided either with raw or processed independent train and test metagenomics data to develop and apply models for classifying metagenomics fecal samples from CD, UC, and non-IBD subjects. The challenge results answer a set of scientific questions among which are: whether metagenomics data are sufficiently informative to predict IBD status, with potential applications for IBD diagnostics; what is the nature of the most discriminative metagenomic features (taxonomy, pathway, other), and are they distinct in UC and CD; and which predictive computational approach is the most accurate for each of the classification tasks."
Short Abstract: Advancements in sequencing technologies has enabled sequencing entire microbiomes present in an environment possible. While initial approaches involved sequencing short-reads to adequate depths- thereby requiring assembly and/or mapping to reference genomes, Oxford nanopore technology (ONT) made it possible to obtain long sequences ranging from thousands, to hundreds of thousands bases. This overcomes limitations associated with short reads e.g. long genomic repeats and complex sequences. Furthermore, it is desirable because of its portability and low operational costs. However, a major limitation is inactive and blocked pores resulting from continuous use. In this study, we explored implications of reusing the flowcell to identify microbiomes present in a sample over time. An environmental sample was sequenced in three intervals, 24 hours apart, and finally a week later to monitor yield, accuracy and quality of the flowcell under similar conditions. With subsequent runs, we observed reduced sequencing capacity and efficiency of the flowcell; leading to reduced detection of low-abundant bacterial communities that were previously identified present in the community. Despite this, the underlying microbial patterns and overall diversity in the abundant species was consistent and reliable to form a descriptive baseline adequate to describe the ecological niche of interest.
Short Abstract: Fast and accurate taxonomic classification of bacteria genomes is a key step in human gut microbiome analysis. Current computational tools rely on sequence-alignment-based methods such as ANI, isDDH and phylogenomic reconstruction. Most tools do not differentiate human gut bacteria from the ones of other environments. Here we propose Xlassify, an alignment-free deep-learning model that is specifically trained to classify human gut bacteria. We downloaded Unified Human Gastrointestinal Genome (UHGG), the most comprehensive human gut bacteria collection including 204,938 genomes from 4,644 species. We first extracted 16S rRNA genes from UHGG genomes and built a 16S-rRNA-based deep-learning model that assigns species at 76.4% accuracy. The performance demonstrated the limited resolution of 16S rRNA, especially for the metagenome-assembled genomes (MAGs). We then built Xlassify by adopting a four-layer Residual Network on 7-mer features extracted from UHGG genomes. Xlassify demonstrated 98% accuracy in 5-fold cross validation and ~90% accuracy on an independent testset of 76 gut bacterial genomes isolated from healthy Chinese individuals. Better than alignment-based methods such as GTDBTk, Xlassify requires only <4GB of memory and reaches thirty-second-per-genome speed on a single CPU.
We communicated with Dr. Steven Leard. We will make Xlassify available at no cost to the community.
