Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ANI”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules

Abstract Maximum diversification of data is a central theme in building generalized and accurate machine learning (ML) models. In chemistry, ML has been used to develop models for predicting molecular properties, for example quantum mechanics (QM) calculated potential energy surfaces and atomic charge models. The ANI-1x and ANI-1ccx ML-based general-purpose potentials for organic molecules were developed through active learning; an automated data diversification process. Here, we describe the ANI-1x and ANI-1ccx data sets. To demonstrate data diversity, we visualize it with a dimensionality reduction scheme, and contrast against existing data sets. The ANI-1x data set contains multiple QM properties from 5 M density functional theory calculations, while the ANI-1ccx data set contains 500 k data points obtained with an accurate CCSD(T)/CBS extrapolation. Approximately 14 million CPU core-hours were expended to generate this data. Multiple QM calculated properties for the chemical elements C, H, N, and O are provided: energies, atomic forces, multipole moments, atomic charges, etc. We provide this data to the community to aid research and development of ML models for chemistry.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2017

ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost

Deep learning is revolutionizing many areas of science and technology, especially image, text, and speech recognition. In this paper, we demonstrate how a deep neural network (NN) trained on quantum mechanical (QM) DFT calculations can learn an accurate and transferable potential for organic molecules. We introduce ANAKIN-ME (Accurate NeurAl networK engINe for Molecular Energies) or ANI for short. ANI is a new method designed with the intent of developing transferable neural network potentials that utilize a highly-modified version of the Behler and Parrinello symmetry functions to build single-atom atomic environment vectors (AEV) as a molecular representation. AEVs provide the ability to train neural networks to data that spans both configurational and conformational space, a feat not previously accomplished on this scale. We utilized ANI to build a potential called ANI-1, which was trained on a subset of the GDB databases with up to 8 heavy atoms in order to predict total energies for organic molecules containing four atom types: H, C, N, and O. To obtain an accelerated but physically relevant sampling of molecular potential surfaces, we also proposed a Normal Mode Sampling (NMS) method for generating molecular conformations. Through a series of case studies, we show that ANI-1 is chemically accurate compared to reference DFT calculations on much larger molecular systems (up to 54 atoms) than those included in the training data set.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Synthesis of Trimethyltriazacyclohexane (Me 3 tach) Sandwich Complexes of Uranium, Neptunium, and Plutonium Triiodides: (Me 3 tach) 2 AnI 3

1,3,5-Trimethyl-1,3,5-triazacyclohexane (Me 3 tach) readily complexes uranium triiodide to form (Me 3 tach) 2 UI 3 . Further, the complex is soluble in THF and arenes and can function as a source of UI 3 to form organometallic U(III) complexes. When dissolved in pyridine (py), (Me3tach)2UI3 forms (Me 3 tach)UI 3 (py) 2 . A related complex with the larger 1,4,7-trimethyl-1,4,7-triazacyclononane (Me 3 tacn) ligand, namely (Me 3 tacn)UI 3 (THF), was synthesized for comparison. Since X-ray quality crystals of (Me 3 tach) 2 UI 3 can be synthesized in high yield even with small-scale reactions, the system is ideal for extension to transuranium elements. Accordingly, the neptunium and plutonium complexes (Me 3 tach) 2 NpI 3 and (Me 3 tach) 2 PuI 3 were synthesized in an analogous manner from NpI 3 (THF) 4 and PuI 3 (THF) 4 , respectively.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2017

ANI-1, A data set of 20 million calculated off-equilibrium conformations for organic molecules

One of the grand challenges in modern theoretical chemistry is designing and implementing approximations that expedite ab initio methods without loss of accuracy. Machine learning (ML) methods are emerging as a powerful approach to constructing various forms of transferable atomistic potentials. They have been successfully applied in a variety of applications in chemistry, biology, catalysis, and solid-state physics. However, these models are heavily dependent on the quality and quantity of data used in their fitting. Fitting highly flexible ML potentials, such as neural networks, comes at a cost: a vast amount of reference data is required to properly train these models. We address this need by providing access to a large computational DFT database, which consists of more than 20 M off equilibrium conformations for 57,462 small organic molecules. We believe it will become a new standard benchmark for comparison of current and future methods in the ML potential community.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Differentially Private Adaptive Noise Injection (DP-ANI) v1.0

Location data is collected from users continuously to understand their mobility patterns. Releasing the user trajectories may compromise user privacy. Therefore, the general practice is to release aggregated location datasets. However, private information may still be inferred from an aggregated version of location trajectories. Differential privacy (DP) protects the query output against inference attacks regardless of background knowledge. This software implements a differential privacy-based privacy model that protects the user's origins and destinations from being inferred from aggregated mobility datasets. This is achieved by injecting Planar Laplace noise to the user origin and destination GPS points. The noisy GPS points are then transformed into a link representation using a link-matching algorithm. Finally, the link trajectories form an aggregated mobility network. The injected noise level is selected using the Sparse Vector Mechanism. This DP selection mechanism considers the link density of the location and the functional category of the localized links. Compared to the different baseline models, including a k-anonymity method, our differential privacy-based aggregation model offers query responses that are close to the raw data in terms of aggregate statistics at both the network and trajectory-levels with maximum 9% deviation from the baseline in terms of network length.

Peisert, Sean [Lawrence Berkeley National Laborato↗
DOE OSTI2017

Draft genome sequence of type strain HBR26T and description of Rhizobium aethiopicum sp. nov.

Rhizobium aethiopicum sp. nov. is a newly proposed species within the genus Rhizobium. This species includes six rhizobial strains; which were isolated from root nodules of the legume plant Phaseolus vulgaris growing in soils of Ethiopia. The species fixes nitrogen effectively in symbiosis with the host plant P. vulgaris, and is composed of aerobic, Gram-negative staining, rod-shaped bacteria. The genome of type strain HBR26 T of R. aethiopicum sp. nov. was one of the rhizobial genomes sequenced as a part of the DOE JGI 2014 Genomic Encyclopedia project designed for soil and plant-associated and newly described type strains. The genome sequence is arranged in 62 scaffolds and consists of 6,557,588 bp length, with a 61% G + C content and 6221 protein-coding and 86 RNAs genes. The genome of HBR26 T contains repABC genes (plasmid replication genes) homologous to the genes found in five differen t Rhizobium etli CFN42 T plasmids, suggesting that HBR26 T may have five additional replicons other than the chromosome. In the genome of HBR26 T , the nodulation genes nodB, nodC, nodS, nodI, nodJ and nodD are located in the same module, and organized in a similar way as nod genes found in the genome of other known common bean-nodulating rhizobial species. nodA gene is found in a different scaffold, but it is also very similar to nodA genes of other bean-nodulating rhizobial strains. Though HBR26 T is distinct on the phylogenetic tree and based on ANI analysis (the highest value 90.2% ANI with CFN42 T ) from other bean-nodulating species, these nod genes and most nitrogen-fixing genes found in the genome of HBR26 T share high identity with the corresponding genes of known bean-nodulating rhizobial species (96-100% identity). This suggests that symbiotic genes might be shared between bean-nodulating rhizobia through horizontal gene transfer. R. aethiopicum sp. nov. was grouped into the genus Rhizobium but was distinct from all recognized species of that genus by phylogenetic analyses of combined sequences of the housekeeping genes recA and glnII. The closest reference type strains for HBR26 T were R. etli CFN42 T (94% similarity of the combined recA and glnII sequences) and Rhizobium bangladeshense BLR175 T (93%). Genomic ANI calculation based on protein-coding genes also revealed that the closest reference strains were R. bangladeshense BLR175 T and R. etli CFN42 T with ANI values 91.8 and 90.2%, respectively. Nevertheless, the ANI values between HBR26 T and BLR175 T or CFN42 T are far lower than the cutoff value of ANI ( > = 96%) between strains in the same species, confirming that HBR26 T belongs to a novel species. Thus, on the basis of phylogenetic, comparative genomic analyses and ANI results, we formally propose the creation of R. aethiopicum sp. nov. with strain HBR26 T (=HAMBI 3550 T =LMG 29711 T ) as the type strain. The genome assembly and annotation data is deposited in the DOE JGI portal and also available at European Nucleotide Archive under accession numbers FMAJ01000001-FMAJ01000062.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2018

Less is more: Sampling chemical space with active learning

We present the development of accurate and transferable machine learning (ML) potentials for predicting molecular energetics is a challenging task. The process of data generation to train such ML potentials is a task neither well understood nor researched in detail. In this work, we present a fully automated approach for the generation of datasets with the intent of training universal ML potentials. It is based on the concept of active learning (AL) via Query by Committee (QBC), which uses the disagreement between an ensemble of ML potentials to infer the reliability of the ensemble’s prediction. QBC allows the presented AL algorithm to automatically sample regions of chemical space where the ML potential fails to accurately predict the potential energy. AL improves the overall fitness of ANAKIN-ME (ANI) deep learning potentials in rigorous test cases by mitigating human biases in deciding what new training data to use. AL also reduces the training set size to a fraction of the data required when using naive random sampling techniques. To provide validation of our AL approach, we develop the COmprehensive Machine-learning Potential (COMP6) benchmark (publicly available on GitHub) which contains a diverse set of organic molecules. Active learning-based ANI potentials outperform the original random sampled ANI-1 potential with only 10% of the data, while the final active learning-based model vastly outperforms ANI-1 on the COMP6 benchmark after training to only 25% of the data. Finally, we show that our proposed AL technique develops a universal ANI potential (ANI-1x) that provides accurate energy and force predictions on the entire COMP6 benchmark. Finally, this universal ML potential achieves a level of accuracy on par with the best ML potentials for single molecules or materials, while remaining applicable to the general class of organic molecules composed of the elements CHNO.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2017

Draft genome sequences of Bradyrhizobium shewense sp. nov. ERR11 T and Bradyrhizobium yuanmingense CCBAU 10071 T

The type strain of the prospective Bradyrhizobium shewense sp. nov. ERR11 T , was isolated from a nodule of the leguminous tree Erythrina brucei native to Ethiopia. The type strain Bradyrhizobium yuanmingense CCBAU 10071 T , was isolated from the nodules of Lespedeza cuneata in Beijing, China. The genomes of ERR11 T and CCBAU 10071 T were sequenced by DOE–JGI and deposited at the DOE–JGI genome portal as well as at the European Nucleotide Archive. The genome of ERR11 T is 9,163,226 bp in length and has 102 scaffolds, containing 8548 protein–coding and 86 RNA genes. The CCBAU 10071 T genome is arranged in 108 scaffolds and consists of 8,201,522 bp long and 7776 protein–coding and 85 RNA genes. Both genomes contain symbiotic genes, which are homologous to the genes found in the complete genome sequence of Bradyrhizobium diazoefficiens USDA110 T . The genes encoding for nodulation and nitrogen fixation in ERR11 T showed high sequence similarity with homologous genes found in the draft genome of peanut–nodulating Bradyrhizobium arachidis LMG 26795 T . The nodulation genes nolYA-nodD2D1YABCSUIJ-nolO-nodZ of ERR11 T and CCBAU 10071 T are organized in a similar way to the homologous genes identified in the genomes of USDA110 T , Bradyrhizobium ottawaense USDA 4 and Bradyrhizobium liaoningense CCBAU 05525. The genomes harbor hupSLCFHK and hypBFDE genes that code the expression of hydrogenase, an enzyme that helps rhizobia to uptake hydrogen released by the N 2 -fixation process and genes encoding denitrification functions napEDABC and norCBQD for nitrate and nitric oxide reduction, respectively. The genome of ERR11 T also contains nosRZDFYLX genes encoding nitrous oxide reductase. Based on multilocus sequence analysis of housekeeping genes, the novel species, which contains eight strains formed a unique group close to the B. ottawaense branch. Genome Average Nucleotide Identity (ANI) calculated between the genome sequences of ERR11 T and closely related sequences revealed that strains belonging to B. ottawaense branch (USDA4 and CCBAU15615), were the closest strains to the strain ERR11 T with 95.2% ANI. Type strain ERR11 T showed the highest DDH predicted value with CCBAU15615 (58.5%), followed by USDA 4 (53.1%). Nevertheless, the ANI and DDH values obtained between ERR11 T and CCBAU 15615 or USDA 4 were below the cutoff values (ANI ≥ 96.5%; DDH ≥ 70%) for strains belonging to the same species, suggesting that ERR11 T is a new species. Therefore, based on the phylogenetic analysis, ANI and DDH values, we formally propose the creation of B. shewense sp. nov. with strain ERR11 T (HAMBI 3532 T =LMG 30162 T ) as the type strain.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Exploring the frontiers of condensed-phase chemistry with a general reactive machine learning potential

Abstract Atomistic simulation has a broad range of applications from drug design to materials discovery. Machine learning interatomic potentials (MLIPs) have become an efficient alternative to computationally expensive ab initio simulations. For this reason, chemistry and materials science would greatly benefit from a general reactive MLIP, that is, an MLIP that is applicable to a broad range of reactive chemistry without the need for refitting. Here we develop a general reactive MLIP (ANI-1xnr) through automated sampling of condensed-phase reactions. ANI-1xnr is then applied to study five distinct systems: carbon solid-phase nucleation, graphene ring formation from acetylene, biofuel additives, combustion of methane and the spontaneous formation of glycine from early earth small molecules. In all studies, ANI-1xnr closely matches experiment (when available) and/or previous studies using traditional model chemistry methods. As such, ANI-1xnr proves to be a highly general reactive MLIP for C, H, N and O elements in the condensed phase, enabling high-throughput in silico reactive chemistry experimentation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2012

Identifying gaps in grid middleware on fast networks with the advanced networking initiative

As of 2012, a number of US Department of Energy (DOE) National Laboratories have access to a 100 Gb/s wide-area network backbone. The ESnet Advanced Networking Initiative (ANI) project is intended to develop a prototype network, based on emerging 100 Gb/s Ethernet technology. The ANI network will support DOE's science research programs. A 100 Gb/s network test bed is a key component of the ANI project. The test bed offers the opportunity for early evaluation of 100Gb/s network infrastructure for supporting the high impact data movement typical of science collaborations and experiments. In order to make effective use of this advanced infrastructure, the applications and middleware currently used by the distributed computing systems of large-scale science need to be adapted and tested within the new environment, with gaps in functionality identified and corrected. As a user of the ANI test bed, Fermilab aims to study the issues related to end-to-end integration and use of 100 Gb/s networks for the event simulation and analysis applications of physics experiments. In this paper we discuss our findings from evaluating existing HEP Physics middleware and application components, including GridFTP, Globus Online, etc. in the high-speed environment. These will include possible recommendations to the system administrators, application and middleware developers on changes that would make production use of the 100 Gb/s networks, including data storage, caching and wide area access.

Dykstra, Dave↗
DOE OSTI2021

Microcoleus (Cyanobacteria) form watershed-wide populations without strong gradients in population structure

The relative importance of separation by distance and by environment to population genetic diversity can be conveniently tested in river networks, where these two drivers are often independently distributed over space. To evaluate the importance of dispersal and environmental conditions in shaping microbial population structures, we performed genome-resolved metagenomic analyses of benthic Microcoleus -dominated cyanobacterial mats collected in the Eel and Russian River networks (California, USA). The 64 Microcoleus genomes were clustered into three species that shared >96.5% average nucleotide identity (ANI). Most mats were dominated by one strain, but minor alleles within mats were often shared, even over large spatial distances (>300 km). Within the most common Microcoleus species, the ANI between the dominant strains within mats decreased with increasing spatial separation. However, over shorter spatial distances (tens of kilometres), mats from different subwatersheds had lower ANI than mats from the same subwatershed, suggesting that at shorter spatial distances environmental differences between subwatersheds in factors like canopy cover, conductivity, and mean annual temperature decreases ANI. Since mats in smaller creeks had similar levels of nucleotide diversity (π) as mats in larger downstream subwatersheds, within-mat genetic diversity does not appear to depend on the downstream accumulation of upstream-derived strains. The four-gamete test and sequence length bias suggest recombination occurs between almost all strains within each species, even between populations separated by large distances or living in different habitats. Overall, our results show that, despite some isolation by distance and environmental conditions, sufficient gene-flow occurs among cyanobacterial strains to prevent either driver from producing distinctive population structures across the watershed.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

CRADA Number NFE-19-07845 with American Nanotechnologies, Inc. (CRADA Final Report)

Cooperative Research and Development Agreement (CRADA) NFE-19-07845 between Oak Ridge National Laboratory (ORNL) and American Nanotechnologies, Inc. (ANI) focused on studying the use of dielectrophoresis as a mechanism to purify nanoparticles, particularly semiconducting carbon nanotubes (CNTs). The successful development of low-cost purification processes is a significant bottleneck in the adoption of semiconducting CNTs in commercial semiconducting devices. ANI is a startup company developing scalable systems for performing nanoparticle dielectrophoresis, which has previously been used only in microfluidic devices. The work done under this CRADA is exploring the use of ANI’s patent resonant dielectrophoresis (rDEP) technology and its ability to purify semiconducting CNTs at large scale. While continued work is on-going, r-DEP has proven to be a viable way to control nanoparticles in bulk dispersions. Optimization of the process is now underway to reach minimum viable product and begin material sales. Additionally, ANI and ORNL continue to collaborate on leveraging this technology to reach down the value chain and create new commercial devices.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2021

Genomic Metrics Applied to Rhizobiales ( Hyphomicrobiales ): Species Reclassification, Identification of Unauthentic Genomes and False Type Strains

Taxonomic decisions within the order Rhizobiales have relied heavily on the interpretations of highly conserved 16S rRNA sequences and DNA–DNA hybridizations (DDH). Currently, bacterial species are defined as including strains that present 95–96% of average nucleotide identity (ANI) and 70% of digital DDH (dDDH). Thus, ANI values from 520 genome sequences of type strains from species of Rhizobiales order were computed. From the resulting 270,400 comparisons, a ≥95% cut-off was used to extract high identity genome clusters through enumerating maximal cliques. Coupling this graph-based approach with dDDH from clusters of interest, it was found that: (i) there are synonymy between Aminobacter lissarensis and Aminobacter carboxidus, Aurantimonas manganoxydans and Aurantimonas coralicida, “Bartonella mastomydis,” and Bartonella elizabethae, Chelativorans oligotrophicus, and Chelativorans multitrophicus, Rhizobium azibense, and Rhizobium gallicum, Rhizobium fabae, and Rhizobium pisi, and Rhodoplanes piscinae and Rhodoplanes serenus; (ii) Chelatobacter heintzii is not a synonym of Aminobacter aminovorans; (iii) “Bartonella vinsonii” subsp. arupensis and “B. vinsonii” subsp. berkhoffii represent members of different species; (iv) the genome accessions GCF_003024615.1 (“Mesorhizobium loti LMG 6125T”), GCF_003024595.1 (“Mesorhizobium plurifarium LMG 11892T”), GCF_003096615.1 (“Methylobacterium organophilum DSM 760T”), and GCF_000373025.1 (“R. gallicum R-602 spT”) are not from the genuine type strains used for the respective species descriptions; and v) “Xanthobacter autotrophicus” Py2 and “Aminobacter aminovorans” KCTC 2477T represent cases of misuse of the term “type strain”. Aminobacter heintzii comb. nov. and the reclassification of Aminobacter ciceronei as A. heintzii is also proposed. To facilitate the downstream analysis of large ANI matrices, we introduce here ProKlust (“Prokaryotic Clusters”), an R package that uses a graph-based approach to obtain, filter, and visualize clusters on identity/similarity matrices, with settable cut-off points and the possibility of multiple matrices entries.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Improving Bond Dissociations of Reactive Machine Learning Potentials through Physics-Constrained Data Augmentation

In the field of computational chemistry, predicting bond dissociation energies (BDEs) presents well-known challenges, particularly due to the multireference character of reactive systems. Many chemical reactions involve configurations where single-reference methods fall short, as the electronic structure can significantly change during bond breaking. As generating training data for partially broken bonds is a challenging task, even state-of-the-art reactive machine learning interatomic potentials (MLIPs) often fail to predict reliable BDEs and smooth dissociation curves. By contrast, simple and inexpensive physics-based models, such as the well-established Morse potential, do not suffer from any such limitations. This work leverages the Morse potential to improve reactive MLIPs by augmenting the training data set with inexpensive Morse data along the dissociation pathways. Further, this physics-constrained data augmentation (PCDA) approach results in MLIPs with smooth bond dissociation curves as well as near coupled-cluster level BDEs, all without requiring any expensive multireference quantum mechanical calculations. A case study for methane combustion demonstrates how the PCDA approach can improve an existing reactive MLIP, namely, ANI-1xnr. In conclusion, not only are the BDEs and bond dissociation curves for all radicals and molecules significantly improved compared to ANI-1xnr but the PCDA-trained MLIP retains the reliability of ANI-1xnr when performing reactive molecular dynamics simulations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

A Frequency Domain Methodology for Quantitative Evaluation of Diffuse Wavefield With Applications to Seismic Imaging

Abstract Ambient Noise Imaging (ANI) of subsurface structures relies on seismic interferometry of diffuse seismic wavefields. However, the lack of effective methods to quantify and identify highly diffuse waves hampers applications of ANI, particularly in evaluating seismic attenuation and monitoring structural changes with high temporal resolution. Conventional ANI approaches require data normalization, which effectively suppresses the non‐diffuse component with large amplitude but also results in significant loss of amplitude and phase information in the continuous seismic records. In this study, we propose a frequency domain method to quantitatively evaluate the degree of diffuseness of seismic wavefields by analyzing their statistical characteristics of modal amplitudes for stationarity and randomness. Tests on synthetic waveform and field nodal records show that the proposed method can effectively distinguish between diffuse and non‐diffuse waveforms for either single‐ or three‐component data. As an application, we identify a 60‐s‐long diffuse coda of a local M 2.2 earthquake recorded by a dense nodal array on the San Jacinto Fault Zone, and successfully extract high‐quality dispersion curve andQ‐value without performing data normalization. These results are consistent with those obtained by conventional methods that assess the correlation between coherency and the Green's function, and by modeling ballistic waves generated by road traffic. Our proposed method can advance the imaging of subsurface velocity and attenuation structures as well as monitoring temporal changes for scientific studies and engineering applications.

Geochemistry & Geophysics↗
DOE OSTI2021

Automated discovery of a robust interatomic potential for aluminum

Machine learning, trained on quantum mechanics (QM) calculations, is a powerful tool for modeling potential energy surfaces. A critical factor is the quality and diversity of the training dataset. Here we present a highly automated approach to dataset construction and demonstrate the method by building a potential for elemental aluminum (ANI-Al). In our active learning scheme, the ML potential under development is used to drive non-equilibrium molecular dynamics simulations with time-varying applied temperatures. Whenever a configuration is reached for which the ML uncertainty is large, new QM data is collected. The ML model is periodically retrained on all available QM data. The final ANI-Al potential makes very accurate predictions of radial distribution function in melt, liquid-solid coexistence curve, and crystal properties such as defect energies and barriers. We perform a 1.3M atom shock simulation and show that ANI-Al force predictions shine in their agreement with new reference DFT calculations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2018

How Much Do rRNA Gene Surveys Underestimate Extant Bacterial Diversity?

The most common practice in studying and cataloguing prokaryotic diversity involves the grouping of sequences into operational taxonomic units (OTUs) at the 97% 16S rRNA gene sequence identity level, often using partial gene sequences, such as PCR-generated amplicons. Due to the high sequence conservation of rRNA genes, organisms belonging to closely related yet distinct species may be grouped under the same OTU. With this being said, it remains unclear how much diversity has been underestimated by this practice. To address this question, we compared the OTUs of genomes defined at the 97% or 98.5% 16S rRNA gene identity level against OTUs of the same genomes defined at the 95% whole-genome average nucleotide identity (ANI), which is a much more accurate proxy for species. Our results show that OTUs resulting from a 98.5% 16S rRNA gene identity cutoff are more accurate than 97% compared to 95% ANI (90.5% versus 89.9% accuracy) but indistinguishable from any other threshold in the 98.29 to 98.78% range. Even with the more stringent thresholds, the 16S rRNA gene-based approach commonly underestimates the number of OTUs by ~12%, on average, compared to the ANI-based approach (~14% underestimation when using the 97% identity threshold). Moreover, the degree of underestimation can become 50% or more for certain taxa, such as the genera Pseudomonas , Burkholderia , Escherichia , Campylobacter , and Citrobacter . These results provide a quantitative view of the degree of underestimation of extant prokaryotic diversity by 16S rRNA gene-defined OTUs and suggest that genomic resolution is often necessary. IMPORTANCE: Species diversity is one of the most fundamental pieces of information for community ecology and conservational biology. Thus, employing accurate proxies for what a species or the unit of diversity is are cornerstones for a large set of microbial ecology and diversity studies. The most common proxies currently used rely on the clustering of 16S rRNA gene sequences at some threshold of nucleotide identity, typically 97% or 98.5%. Here, we explore how well this strategy reflects the more accurate whole-genome-based proxies and determine the frequency with which the high conservation of 16S rRNA sequences masks substantial species-level diversity.

16S rRNA gene↗