Search

Oceanic bacterial communities process a major fraction of marine organic carbon. A substantial portion of this carbon transformation occurs in the mesopelagic zone, and a further fraction fuels bacteria in the bathypelagic zone. However, the capabilities and limitations of the diverse microbial communities at these depths to degrade high-molecular-weight (HMW) organic matter are not well constrained. Here, we compared the responses of distinct microbial communities from North Atlantic epipelagic (0–200 m), mesopelagic (200–1000 m), and bathypelagic (1000–4000 m) waters at two open-ocean stations to the same input of diatom-derived HMW particulate and dissolved organic matter. Microbial community composition and functional responses to the input of HMW organic matter – as measured by polysaccharide hydrolase, glucosidase, and peptidase activities – were very similar between the stations, which were separated by 1370 km but showed distinct patterns with depth. Changes in microbial community composition coincided with changes in enzymatic activities: as bacterial community composition changed in response to the addition of HMW organic matter, the rate and spectrum of enzymatic activities increased. In epipelagic mesocosms, the spectrum of peptidase activities became especially broad and glucosidase activities were very high, a pattern not seen at other depths, which, in contrast, were dominated by leucine aminopeptidase and had much lower peptidase and glucosidase rates in general. The spectrum of polysaccharide hydrolase activities was enhanced particularly in epipelagic and mesopelagic mesocosms, with fewer enhancements in rates or spectrum in bathypelagic waters. The timing and magnitude of these distinct functional responses to the same HMW organic matter varied with depth. Our results highlight the importance of residence times at specific depths in determining the nature and quantity of organic matter reaching the deep sea.

Spaceflight has several detrimental effects on human and rodent health. For example, liver dysfunction is a common phenotype observed in space-flown rodents, and this dysfunction is partially reflected in transcriptomic changes. Studies linking transcriptomics with liver dysfunction rely on tools which exploit correlation, but these tools make no attempt to disambiguate true correlations from spurious ones. In this work, we use a machine learning ensemble of causal inference methods called the Causal Research and Inference Search Platform (CRISP). It was specifically developed to predict causal features of a binary response variable in high-dimensional input. We used CRISP to search gene expression data for genes truly correlated with a lipid density phenotype using transcriptomic and histological data from the NASA Space Biology Open Science Data Repositories (OSDR). Our approach identified genes and molecular targets not predicted by previous traditional system biology approaches. These genes are likely to play a pivotal role in the liver dysfunction observed in space-flown rodents, and this work opens the door to identifying novel countermeasures for space travel.

We introduce CRISP, a Causal Research and Inference Search Platform. It is designed to assist biological and medical research by applying a variety of causal discovery methods to heterogeneous and high-dimensional observational data. CRISP aims to identify a small set of input variables which are most likely to have a causal effect on a target variable. The output of CRISP, thus, highlights the most promising candidates for further targeted research. We illustrate the utility of CRISP with a case study in oncology, using a multi-omic colorectal cancer data set to identify causal drivers differentiating two subtypes of colorectal cancer.

The majority of microbial genomes have yet to be cultured, and most proteins predicted from microbial genomes or sequenced from the environment cannot be functionally annotated. As a result, current computational approaches to describe microbial systems rely on incomplete reference databases that cannot adequately capture the full functional diversity of the microbial tree of life, limiting our ability to model high-level features of biological sequences. The scientific community needs a means to capture the functionally and evolutionarily relevant features underlying biology, independent of our incomplete reference databases. Such a model can form the basis for transfer learning tasks, enabling downstream applications in environmental microbiology, medicine, and bioengineering. Here we present LookingGlass, a deep learning model capturing a “universal language of life”. LookingGlass encodes contextually-aware, functionally and evolutionarily relevant representations of short DNA reads, distinguishing reads of disparate function, homology, and environmental origin. We demonstrate the ability of LookingGlass to be fine-tuned to perform a range of diverse tasks: to identify novel oxidoreductases, to predict enzyme optimal temperature, and to recognize the reading frames of DNA sequence fragments. LookingGlass is the first contextually-aware, general purpose pre-trained “biological language” representation model for short-read DNA sequences. LookingGlass enables functionally relevant representations of otherwise unknown and unannotated sequences, shedding light on the microbial dark matter that dominates life on Earth.AvailabilityThe pretrained LookingGlass model and the transfer learning-derived models demonstrated in this paper are available in the LookingGlass release v1.01. The open source fastBio Github repository and python package provides classes and functions for training and fine tuning deep learning models with biological data2. Code for reproducing analyses presented in this paper are available as an open source Github repository3.

Background The number of applications of deep learning algorithms in bioinformatics is increasing as they usually achieve superior performance over classical approaches, especially, when bigger training datasets are available. In deep learning applications, discrete data, e.g. words or n-grams in language, or amino acids or nucleotides in bioinformatics, are generally represented as a continuous vector through an embedding matrix. Recently, learning this embedding matrix directly from the data as part of the continuous iteration of the model to optimize the target prediction – a process called ‘end-to-end learning’ – has led to state-of-the-art results in many fields. Although usage of embeddings is well described in the bioinformatics literature, the potential of end-to-end learning for single amino acids, as compared to more classical manually-curated encoding strategies, has not been systematically addressed. To this end, we compared classical encoding matrices, namely one-hot, VHSE8 and BLOSUM62, to end-to-end learning of amino acid embeddings for two different prediction tasks using three widely used architectures, namely recurrent neural networks (RNN), convolutional neural networks (CNN), and the hybrid CNN-RNN. Results By using different deep learning architectures, we show that end-to-end learning is on par with classical encodings for embeddings of the same dimension even when limited training data is available, and might allow for a reduction in the embedding dimension without performance loss, which is critical when deploying the models to devices with limited computational capacities. We found that the embedding dimension is a major factor in controlling the model performance. Surprisingly, we observed that deep learning models are capable of learning from random vectors of appropriate dimension. Conclusion Our study shows that end-to-end learning is a flexible and powerful method for amino acid encoding. Further, due to the flexibility of deep learning systems, amino acid encoding schemes should be benchmarked against random vectors of the same dimension to disentangle the information content provided by the encoding scheme from the distinguishability effect provided by the scheme.

SAR86 is an abundant and ubiquitous heterotroph in the surface ocean that plays a central role in the function of marine ecosystems. We hypothesized that despite its ubiquity, different SAR86 subgroups may be endemic to specific ocean regions and functionally specialized for unique marine environments. However, the global biogeographical distributions of SAR86 genes, and the manner in which these distributions correlate with marine environments, have not been investigated. We quantified SAR86 gene content across globally-distributed metagenomic samples and modeled these gene distributions as a function of 51 environmental variables. We identified five distinct clusters of genes within the SAR86 pangenome, each with a unique geographic distribution associated with specific environmental characteristics. Gene clusters are characterized by strong taxonomic enrichment of distinct SAR86 genomes and partial assemblies, as well as differential enrichment of certain functional groups, suggesting differing functional and ecological roles of SAR86 ecotypes. We then leveraged our models and high-resolution, remote sensing-derived environmental data to predict the distributions of SAR86 gene clusters across the world’s oceans, creating global maps of SAR86 ecotype distributions. Our results reveal that SAR86 exhibits previously unknown, complex biogeography, and provide a framework for exploring geographic distributions of genetic diversity from other microbial clades.

Organic carbon in marine sediments is a critical component of the global carbon cycle, and its degradation influences a wide range of phenomena, including the magnitude of carbon sequestration over geologic timescales, the recycling of inorganic carbon and nutrients, the dissolution and precipitation of carbonates, the production of methane and the nature of the seafloor biosphere. Although much has been learned about the factors that promote and hinder rates of organic carbon degradation in natural systems, the controls on the distribution of organic carbon in modern and ancient sediments are still not fully understood. In this review, we summarize how recent findings are changing entrenched perspectives on organic matter degradation in marine sediments: a shift from a structurally-based chemical reactivity viewpoint towards an emerging acceptance of the role of the ecosystem in organic matter degradation rates. That is, organic carbon has a range of reactivities determined by not only the nature of the organic compounds, but by the biological, geochemical, and physical attributes of its environment. This shift in mindset has gradually come about due to a greater diversity of sample sites, the molecular revolution in biology, discoveries concerning the extent and limits of life, advances in quantitative modeling, investigations of ocean carbon cycling under a variety of extreme paleo-conditions (e.g. greenhouse environments, euxinic/anoxic oceans), the application of novel analytical techniques and interdisciplinary efforts. Adopting this view across scientific disciplines will enable additional progress in understanding how marine sediments influence the global carbon cycle.

Warm core, anticyclonic rings that spin off from the Gulf Stream circulate through the region directly offshore of the Mid-Atlantic Bight. If a warm core ring reaches the continental shelf break, its warm, highly saline water may subduct under cooler, fresher continental shelf surface water, resulting in subsurface waters at the shelf break and over the upper continental slope with high temperatures and salinities and distinct physical and chemical properties characteristic of Gulf Stream water. Such intruding water may also have microbial communities with distinct functional capacities, which may in turn affect the rate and nature of carbon cycling in this coastal/shelf environment. However, the functional capabilities of microbial communities within ring intrusion waters relative to surrounding continental shelf waters are largely unexplored. We investigated microbial community capacity to initiate organic matter remineralization by measuring hydrolysis of a suite of polysaccharide, peptide, and glucose substrates along a transect oriented across the Mid-Atlantic Bight shelf, shelf break, and upper slope. At the outermost sampling site, warm and salty water derived from a Gulf Stream warm core ring was present in the lower portion of the water column. This water exhibited hydrolytic capacities distinct from other sampling sites, and exhibited lower heterotrophic bacterial productivity overall. Warm core rings adjacent to the Mid-Atlantic Bight shelf have increased in frequency and duration in recent years. As the influence of warm core rings on the continental shelf and slope increases in the future, the rate and nature of organic matter remineralization on the continental shelf may also shift.

The extent to which differences in microbial community structure result in variations in organic matter (OM) degradation is not well understood. Here, we tested the hypothesis that distinct marine microbial communities from North Atlantic surface and bottom waters would exhibit varying compositional succession and functional shifts in response to the same pool of complex high molecular weight (HMW-OM). We also hypothesized that microbial communities would produce a broader spectrum of enzymes upon exposure to HMW-OM, indicating a greater potential to degrade these compounds than reflected by initial enzymatic activities. Our results show that community succession in amended mesocosms was congruent with cell growth, increased bacterial production and most notably, with substantial shifts in enzymatic activities. In all amended mesocosms, closely related taxa that were initially rare became dominant at time frames during which a broader spectrum of active enzymes were detected compared to initial timepoints, indicating a similar response among different communities. However, succession on the whole-community level, and the rates, spectra and progression of enzymatic activities, reveal robust differences among distinct communities from discrete water masses. These results underscore the crucial role of rare bacterial taxa in ocean carbon cycling and the importance of bacterial community structure for HMW-OM degradation.

Sequencing data resources have increased exponentially in recent years, as has interest in large-scale meta-analyses of integrated next-generation sequencing datasets. However, curation of integrated datasets that match a user’s particular research priorities is currently a time-intensive and imprecise task. MetaSeek is a sequencing data discovery tool that enables users to flexibly search and filter on any metadata field to quickly find the sequencing datasets that meet their needs. MetaSeek automatically scrapes metadata from all publicly available datasets in the Sequence Read Archive, cleans and parses messy, user-provided metadata into a structured, standard-compliant database, and predicts missing fields where possible. MetaSeek provides a web-based graphical user interface and interactive visualization dashboard, as well as a programmatic API to rapidly search, filter, visualize, save, share, and download matching sequencing metadata. The MetaSeek online interface is available at https://www.metaseek.cloud/. The MetaSeek database can also be accessed via API to programmatically search, filter, and download all metadata. MetaSeek source code, metadata scrapers, and documents are available at https://github.com/MetaSeek-Sequencing-Data-Discovery/metaseek/. Additional guides, tutorials, and documents are available at https://github.com/MetaSeek-Sequencing-Data-Discovery/metaseek, and on the MetaSeek website, https://www.metaseek.cloud/. MetaSeek is distributed under an MIT license.

Extracellular enzymatic activities initiate microbially-driven heterotrophic carbon cycling in subsurface sediments. While measurement of hydrolytic activities in sediments is fundamental to our understanding of carbon cycling, these measurements are often technically difficult due to sorption of organic substrates to the sediment matrix. Most methods that measure hydrolysis of organic substrates in sediments rely on recovery of a fluorophore or fluorescently-labeled target substrate from a sediment incubation. The tendency for substrates to sorb to sediments results in lower recovery of an added substrate, and can result in data that are unusable or difficult to interpret. We developed a treatment using competitive desorption of a fluorescently-labeled, high molecular weight organic substrate that improves recovery of the labeled substrate from sediment subsamples. Competitive desorption treatment improved recovery of the fluorescent substrate by a median of 66%, expanded the range of sediments for which activity measurements could be made, and was effective in sediments from a broad range of geochemical contexts. More reliable measurements of hydrolytic activities in sediments will yield usable and more easily interpretable data from a wider range of sedimentary environments, enabling better understanding of microbially-catalyzed carbon cycling in subsurface environments.