Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Curation”

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 1,063 records · Page 59

Enumeration and Fluorescence In Situ Hybridization of Microbial Bioburden on Cleanroom Surfaces

Introduction: Microorganisms are everywhere on Earth, even in the cleanest of places. Spacecraft assembly cleanrooms can harbor low levels of living and dead microbial cells (e.g., [1,2]), and cleanroom bioburden can also include organic molecules from industrial sources and in situ biomass. Life detection missions require careful attention to avoid contaminants that can be easily convoluted with analytical targets. We are evaluating epifluorescent microscopy and fluorescence in situ hybridization (FISH) as methods to complement organic contamination detection techniques. Epifluorescent cell counting offers an accurate and cost-effective way to quantify low levels of surface biomass. FISH could allow for the identification of residual organisms, and can be targeted to detect active populations of specific organisms such as bacteria known to resist cleaning procedures. This effort is part of a larger study that is concentrated on characterizing the surface and airborne molecular organic contamination background in Johnson Space Center (JSC) Astromaterials curation laboratories and Goddard Space Flight Center (GSFC) spacecraft assembly rooms, and understanding contaminants in the context of cleaning procedures and residual bioburden. Methods: Samples were collected by swabbing surfaces in ISO 5 and ISO 7 equivalent cleanrooms at JSC. Swabs for FISH were fixed in 4% paraformaldehyde (PFA) for 3 hours and then stored in 1:1 ethanol:PBS, while swabs for cell counting were stored in 4% PFA until analysis to avoid any cell loss during centrifugation that could impact quantification of very low biomass samples. Cell counting was performed with SYBR Gold as in [3], but adapted for very low biomass. FISH was performed as in [4], using DAPI as a counterstain for all DNA-containing cells. Negative controls included wells with no probe applied, to test for natural fluorescence, as well as the nonsense probe NONEUB (reverse complement of EUB338) to evaluate non-specific probe binding. Results and Discussion: Cleanroom surfaces had 102-103 cells cm-2. The extremely low biomass of these samples was challenging for enumeration, and required careful and routine use of “field” and laboratory blanks. FISH was performed with the general archaeal and bacterial probes ARCH915 and EUB338 (EUBMIX, [4]), probe GAMBET ([4]), and PSE227, which targets the genus Pseudomonas [5]). The latter two probes were selected because Pseudomonas spp. and other Gammaproteobacteria have not been isolated from cleanroom surfaces but do appear frequently in rRNA gene libraries from these surfaces. While some active bacteria were identified (Fig. 1c), most cells detectable by DAPI did not have a strong or any fluorescent signal (e.g., Fig. 1d), indicating that the vast majority of cells are dead or inactive. This suggests that cleaning protocols are effective at inactivating microbial contaminants, but that dead or inactive cells can remain on surfaces. Cells were often clumped in a weakly autofluorescent matrix, possibly biofilm material (Fig. 1c,d). We also observed other particulate material that was collected by the swabs, including apparent textile fibers (Fig. 1b). Our results are consistent with other studies that show that the bioburden present in clean rooms includes active, dormant, and dead cells. We will discuss how FISH and epifluorescent cell counting could be applied in planetary protection protocols, including the advantages and disadvantages of FISH and cell counting for routine use, as well as different possible applications for more specialized FISH procedures. References: [1] Moissl-Eichinger et al. (2015) Sci Rep, 5, 9156 [2] Hendrickson et al. (2021) Microbiome, 9, 238 [3] Jones et al. (2017) Appl Environ Microbiol, 83, e00909-17 [4] Jones et al. (2015) Appl Environ Microbiol, 81, 1242-1250. [5] Watt et al. (2006) Environ Microbiol, 8, 871-884

C J Huff↗

Maine Ecological Forecasting III: Utilizing Earth Observations to Monitor Federally Endangered Atlantic Salmon (Salmo salar) Habitat in Maine: An Interactive Workshop

Shifting patterns in land use and land cover (LULC), temperature, and precipitation have exacerbated a rapid decline in Federally Endangered wild Atlantic salmon (Salmo salar) populations. The team at NASA DEVELOP partnered with the Maine Department of Marine Resources (DMR) and the Downeast Salmon Federation (DSF) to create a comprehensive workshop designed to demonstrate the applicability of Earth observations in examining these threats using the Penobscot, Union, and Machias Rivers as case studies. This entailed curating tutorials for acquiring and analyzing satellite data using Google Earth Engine, EarthExplorer, and Earthdata. The team demonstrated how to classify LULC in ArcGIS Pro from 1985 until 2021 using Landsat 5 Thematic Mapper (TM), Landsat 8 Operation Land Imager (OLI), Sentinel-2 MultiSpectral Instrument (MSI), and datasets from the United Stated Geological Survey (USGS) National Land Cover Database (NLCD), showing an overall transition from coniferous forests to other LULC classes. The team also demonstrated how to use historical data from Terra Moderate Resolution Imaging Spectroradiometer (MODIS) and Integrated Multi-satellite Retrievals for Global Precipitation Measurement (GPM IMERG) to generate 2021 land surface temperature (LST) and precipitation maps, respectively, showing that Maine was abnormally dry during the summer in an increasingly warm region. These workshop materials will aid the partners in integrating NASA Earth observations into their future salmon habitat restoration initiatives.

Jonathan Falciani↗

NASA Life Sciences Portal (NLSP): Supporting Scientific Transparency and Reproducibility

NASA’s Life Sciences Ports (NLSP) serves the scientific community by providing curated data from space life science experiment. The Human Research Program (HRP) with the help of NLSP is currently transforming their life sciences data archive systems and processes to improve compliance with the FAIR principles [1]. Some of these improvements will at the same time support the twin pillars of Open Science [2]: transparency of methods and reproducibility of results. Scientific transparency is marked by the easily intelligible communication of what has been investigated: what were the procedures for collecting sample and the characteristics of samples collected? what kinds of measurements were made, what were the environmental conditions of the measurements? What were the analysis techniques of the collected data? Reproducibility of the results and findings from the investigation requires a high level of transparency for all but the simplest investigations; the slightest deviation in communicating and replicating complex experimental procedures or data analyses can often yield quite different data and even findings, thwarting their validation. One of the ways the NLSP is aiming to improve the communication of scientific information is through the use of ontology-driven metadata. Ontologies are powerful, graph-based knowledge representation structures, which can be leveraged to increase data interoperability, the area of the FAIR principles in which many data systems most lack compliance. Over the past decade, there has been a concerted effort in the biomedical community to develop modular and narrowly focused domain and application-specific ontologies in a common, open-source framework, the Open Biological and Biomedical Ontology (OBO) Foundry [3]. The open sharing and modular nature of this effort promises huge increases in harmonized data sharing for systems that leverage these models. Which is in line with the FAIR Data Principles of Findability, Accessibility, Interoperability, and Reuse for scientific data management and stewardship. 1. Wilkinson, M.D., et al., The FAIR Guiding Principles for scientific data management and stewardship. Sci Data, 2016. 3: p. 160018. 2. National Academies of Sciences, E. and Medicine, Open Science by Design: Realizing a Vision for 21st Century Research. 2018, Washington, DC: The National Academies Press. 232. 3. Smith, B., et al., The OBO Foundry: coordinated evolution of ontologies to support biomedical data integration. Nat Biotechnol, 2007. 25(11): p. 1251-5.

Life Sciences data↗

Lithological variation of asteroid Ryugu samples returned by the Hayabusa 2 spacecraft: Assessment from the 18 particles distributed to the initial analysis “Stone” team

JAXA’s Hayabusa2 spacecraft successfully returned~5.4g of C-type asteroid Ryugu materials on Dec. 2020 [e.g., 1] and the recovered samples were extensively analyzed by initial analysis teams and Phase2 curation teams. The reported results show that Ryugu samples are similar to CI chondritesin chemistry and mineralogy, showing evidence for aqueous alteration in the parent body [e.g., 2,3]. As a “stone” team of the initial analysis, we received 18coarse particles (>1 mm) of Ryugu samples to characterize their mineralogy and petrology. We found that the samples were breccias of mm-to-sub mm size clasts. All of the clasts are mainly composed of Mg-Fe phyllosilicates, but the alteration degree appears slightly different from one clast to another. Here we report a lithological variation of Ryugu samples to propose their reasonable lithological classification based upon different mineral assemblages and to discuss the formation and evolution of the Ryugu parent body.

Michael E Zolensky↗

Spaceflight Environmental-Telemetry Data for Biological Science

There is a critical need for better access and visualization of spaceflight environmental telemetry and mission hardware data from sensors including relative humidity, carbon dioxide, oxygen, radiation, airflow, temperature, acceleration, and acoustics. Under the stewardship of the Ames Life Sciences Data Archive (ALSDA) and GeneLab, an effort is underway to consolidate, normalize and provide accessibility of archived mission environmental data and hardware information, with the purpose of providing important context to biological data. This effort is necessary to provide scientific context of its impact upon biological and biomedical data from spaceflight missions and experiments (genomic, metagenomic, gene expression, proteomic, metabolomic, physiological, phenomics, behavioral; tabular, imaging, video). Environmental spaceflight data is derived from dozens of sources, with various formats, and in the past year a pipeline is in development to collect, curate and present this data efficiently. In the upcoming year, a new Data Visualization Portal will utilize the standardized pipeline data to provide easy user access to compare parameters and environmental conditions between missions, locations, subjects, and durations. Environmental and hardware data enables broad accessibility and analytics, without the need for advanced data informatic expertise. Familiarity with the capabilities and limitations of a variety of existing hardware/tools is a strength that could be applied to creation of improved hardware for future ecosystems on the Moon and Mars. The intention is to make biological and environmental telemetry data maximally open-access and FAIR (findable, accessible, interoperable, reusable) for data mining-informatic approaches to support knowledge discovery necessary for low Earth orbit, cis-Lunar, Mars transit, and Mars surface missions.

Danielle K. Lopez↗

Contamination monitoring of the OSIRIS-REx ISO5 asteroid sample cleanroom

The OSIRIS-REx mission to asteroid Bennu successfully collected hundreds of grams of asteroid regolith in October 2020 [1]. The spacecraft departed Bennu in May 2021 for Earth return and will release the sample canister to be recovered in Utah in September 2023 [2]. Samples will be transported to the curation facility at NASA Johnson Space Center, where an ISO5 equivalent cleanroom has been designed in 2017 and completed in 2021. Aspects of the design and material selection for the cleanroom and its supporting facilities (air handling system, cleanroom floors and walls, filters, paints, etc.) were optimized to minimize effects of organic and inorganic contaminants and offgassing [3]. Since its completion, the lab has been carefully monitored to understand and establish a baseline with respect to multiple environmental aspects – measurement of particle counts, deploying Si wafer witness plates for organic and inorganic contaminants, deploying aluminum foils for a focus on organics with JSC in-house expertise, gas samples, and regular microbial and fungal measurements on selected surfaces and air samples in the cleanroom. This contribution will report on nearly one year of monitoring and highlight several specific aspects that have led to a better understanding of the new cleanroom environment.

carbonaceous asteroid↗

Safety Culture at the World’s Premier Multi-User Spaceport

NASA’s Agency-wide Safety Culture is implemented at the Kennedy Space Center (KSC) using a unique strategy due an unparalleled approach in making human spaceflight history. Kennedy Space Center, the world’s premier multi-user spaceport, enables U.S. government and commercial space access, while providing the world a resource to allow the exploration of and the ability to work in space. A consistently healthy safety culture at KSC is imperative for mission success: desired achievements, protection of space flight hardware, and ultimately, the preservation of human life requires the support of a healthy safety culture. Emphasis on the NASA Agency-wide development of Safety Culture began with the conception of the NASA Agency Safety Culture Working Group. After the devastating loss of life and mission of the Space Shuttle Colombia, a broken safety culture was identified as an organizational cause by the Columbia Accident Investigation Board Report. Thus, the Agency Safety Culture Working Group was developed in 2009 to assess the status of the Agency’s Safety Culture, while addressing safety culture concerns at the NASA Center-level. A Five-factor model was developed to serve as the guiding principles for Safety Culture: 1) Reporting Culture, 2) Just Culture, 3) Flexible Culture, 4) Learning Culture, and 5) Engaged Culture. These five factors are included in the NASA Safety Culture logo, which was intentionally designed as a DNA double helix to prompt the permeation of safety into day-to-day work. KSC specifically implements the NASA Agency-wide Safety Culture principles in a tailored approach that is relevant to the diversity of work being performed. An emphasis is placed on the implementation to include safety at home, not exclusively at work. This emphasis is a KSC-specific element that has been intentionally added to promote a closed loop Safety Culture. To advertise the safety culture, various safety and health events are held throughout the calendar year, providing innovative speakers and engaging activities, while also promoting a wide range of curated safety initiatives. Development and continuous improvement of the KSC safety tracking database, allows for advanced tracking-to-closure, along with providing data sets used to identify areas of emphasis. Other safety initiatives rely solely on employee participation, such as the photo challenges; participants are encouraged to identify and capture themselves, coworkers, or family members participating in safe or healthy activities to share with others within the Center and at Agency levels. Fabrication of exclusive videos and graphics are utilized to advertise and inform employee of safety initiatives, upcoming safety events, and general dispersion of safety information. In addition, an anonymous Agency-wide Safety Culture Survey is advertised, administered, and analyzed at Kennedy Space Center, with the purpose of receiving basic feedback on Safety Culture perceptions to help prevent future incidents from occurring. Through these briefly identified means, and many other forms of employee engagement, Kennedy Space Center aims to maintain safety in the forefront, while creating an environment where everyone trusts that safety is a priority.

Larrin E. Moody↗

Enabling Space Biological Knowledge Discovery Through Image and Video Data Sharing

Increased biomedical risks associated with deep space crewed missions (cis-Lunar, Mars transit/surface) require development of health countermeasures, novel ecosystem support, risk modeling, and fundamental space biological knowledge discovery. Molecular-omics, physiological-phenotypic-behavioral, and environmental-radiation telemetry data from space biological and health studies are needed for reuse by scientists to address these tasks. The data as well as space-relevant biospecimens are being made more findable, accessible, interoperable, and reusable through NASA’s Open Science Data Repository (OSDR). This new OSDR umbrella grouping includes NASA GeneLab, the NASA Ames Life Sciences Data Archive (ALSDA), and the NASA Biological Institutional Scientific Collection. The OSDR system design appropriately handles metadata and processed-tabular results from ALSDA studies collected from space experiments. But raw and processed ALSDA bioimage and video datasets require an expansion of OSDR’s data architecture to handle ingestion, curation, and egress. The academic-industry bioimaging field saw a scientific renaissance in the past several years through leveraging open-source software, international collaborations, machine learning, and other open science/programming approaches. As crewed missions and more biological experiments are on the deep space horizon, OSDR is embracing data stewardship through listening to feedback from subject matter experts and designing an expanded architecture which is appropriate for NASA’s goals to enable analysis and reuse of bioimaging and video data for the public science community.Discovery Through Image and Video Data Sharing

space biology↗

Making Better Use of Satellite Data: The Satellite Needs Working Group

The U.S. Group on Earth Observations (USGEO) initiated in 2016 the Satellite Needs Working Group (SNWG) to identify and communicate the Earth observation needs of U.S. federal agencies. The SNWG identifies such needs through a biennial survey followed by interviews and follow-up discussions by the satellite Earth data providers of the U.S. Government: the National Aeronautics and Space Administration (NASA), the National Oceanic and Atmospheric Administration (NOAA), and the U.S. Geological Survey (USGS). Solutions and services are identified that leverage current or upcoming satellite missions to meet the identified needs; implementation of services that are estimated to significantly increase the level of satisfaction of multiple U.S. agencies are funded by NASA. The SNWG process has resulted in the implementation of numerous services that have impacted operations of not only U.S. agencies but of academic and international institutions as well. Notable examples include the Harmonized Landsat Sentinel-2 project that leverages European Space Agency (ESA) and NASA satellite assets to generate a global, analysis-ready, surface reflectance product with a temporal resolution of two days; the Airborne Data Management Group that curates and provides access to relevant resources, information and data from existing and past NASA field campaigns; and the Dynamic Surface Water Extent product which consists of harmonized but independent water extents derived from both from optical and radar data. These and other products are hosted at the NASA Distributed Active Archive Centers (DAACs) for free and open access. The SNWG Management Office at NASA’s Interagency Implementation and Advanced Concepts Team (IMPACT) manages the implementation of selected solutions and, importantly, NASA’s response to the needs of federal agencies through a Stakeholder Engagement Program. The impact of implemented services and solutions is hampered without efforts to build capacity around the use of the services. The Stakeholder Engagement Program ensures relevant training and outreach to SNWG agencies in collaboration with each solution implementation team.

remote sensing↗

Resolution of Maldi-Tof Compared to Whole Genome Sequencing for Identification of Bacillus Species Isolated From Cleanrooms at Nasa Johnson Space Center

The Astromaterials Acquisition and Curation Office at NASA Johnson Space Center maintains cleanrooms to archive extraterrestrial materials returned from space exploration missions. Compared to typical built environments, oligotrophic conditions make these facilities inhospitable to microbes. Despite these controls, bacteria and fungi are regularly cultured from these cleanrooms. In particular, Bacillus sp. are frequently isolated during routine microbial monitoring. Endospores associated with this genus can survive extreme environments, such as cleanrooms. This microbial contamination may affect the integrity of astromaterials.

Microbiology↗

Sample Return and Preliminary Examination Timeline for the OSIRIS-REx Mission

NASA’s OSIRIS-REx spacecraft collected a sample of asteroid (101955) Bennu on October 20, 2020, using the Touch-and-Go Sample Acquisition Mechanism (TAGSAM) [1,2,3]. After sample acquisition, the TAGSAM collection head was stowed in the Sample Return Capsule (SRC; Fig. 1) for transport back to Earth. On September 24, 2023, the OSIRIS-Rex spacecraft will fly by Earth and release the SRC, which will land at the Utah Test and Training Range (UTTR). The sample will be transported to the OSIRIS-REx curation facility, located within NASA’s Johnson Space Center (JSC), where it will be examined, cataloged, and allocated for scientific analyses. Here we describe the timeline of operations for the return, recovery, and initial characterization of the Bennu sample.

Asteroid↗

Multi-Decadal Nitrogen Dioxide and Derived Products from Satellites (MINDS) Datasets Released by NASA GES DISC and Their Applications for Air Quality

Nitrogen dioxide (NO2), a pervasive air pollutant, comes from vehicles, power plants, industrial emissions, and off-road sources such as construction or lawn and gardening equipment. The NASA Goddard Earth Sciences Data and Information Services Center (GES DISC) curates many remote sensing datasets with NO2 retrievals, which have been utilized for air quality research and applications. The remotely-sensed datasets include those generated by the Ozone Monitoring Instrument (OMI) on the Aura satellite, the TROPOspheric Monitoring Instrument (TROPOMI) onboard the Copernicus Sentinel-5 Precursor (S5P), and the Ozone Mapping and Profiling Suite (OMPS) Nadir-Mapper (NM) instrument on the Suomi National Polar-orbiting Partnership (S- NPP). In collaboration with the NASA Making Earth System Data Records for Use in Research Environments (MEaSUREs) Multi-Decadal Nitrogen Dioxide and Derived Products from Satellites (MINDS) project, the GES DISC recently released MINDS datasets. The NASA MEaSUREs MINDS project aims to develop long-term NO2 global data records by adapting a consistent retrieval algorithm to multiple instrument measurements. Long-term data records will be achieved by applying consistent retrieval approaches to multiple satellite instruments, including OMI (2004 - ); the Global Ozone Monitoring Experiment (GOME, 1995-2011) onboard the second European Remote Sensing satellite (ERS-2); the Scanning Imaging Spectrometer for Atmospheric Cartography (SCIAMACHY, 2002-2012) onboard the ENVIronmental SATellite (ENVISAT); GOME-2 on the Meteorological Operational satellites (MetOp-A and MetOp-B, 2006 - ); and TROPOMI onboard the Copernicus S5P (2017 - ). The long-term record (1995 to present) of MINDS datasets makes them very useful for air quality trend studies. Some MINDS datasets with high spatial resolution of only a few kilometers can be used for air quality research and applications at regional scales. In this presentation, we will introduce all of the MINDS products and services, and demonstrate use cases of MINDS data for studying air quality. We will also present a few other NO2 datasets acquired from NASA’s Health and Air Quality Applied Sciences Team (HAQAST), to be archived and distributed by the GES DISC, and highlight some of their applications for air quality and health.

Feng Ding↗

NASA Life Sciences Portal (NLSP): Supporting Scientific Transparency and Reproducibility

NASA’s Life Sciences Ports (NLSP) serves the scientific community by providing curated data from space life science experiment. The Human Research Program (HRP) with the help of NLSP is currently transforming their life sciences data archive systems and processes to improve compliance with the FAIR principles [1]. Some of these improvements will at the same time support the twin pillars of Open Science [2]: transparency of methods and reproducibility of results. Scientific transparency is marked by the easily intelligible communication of what has been investigated: what were the procedures for collecting sample and the characteristics of samples collected? what kinds of measurements were made, what were the environmental conditions of the measurements? What were the analysis techniques of the collected data? Reproducibility of the results and findings from the investigation requires a high level of transparency for all but the simplest investigations; the slightest deviation in communicating and replicating complex experimental procedures or data analyses can often yield quite different data and even findings, thwarting their validation. One of the ways the NLSP is aiming to improve the communication of scientific information is through the use of ontology-driven metadata. Ontologies are powerful, graph-based knowledge representation structures, which can be leveraged to increase data interoperability, the area of the FAIR principles in which many data systems most lack compliance. Over the past decade, there has been a concerted effort in the biomedical community to develop modular and narrowly focused domain and application-specific ontologies in a common, open-source framework, the Open Biological and Biomedical Ontology (OBO) Foundry [3]. The open sharing and modular nature of this effort promises huge increases in harmonized data sharing for systems that leverage these models. Which is in line with the FAIR Data Principles of Findability, Accessibility, Interoperability, and Reuse for scientific data management and stewardship.

Life Sciences data↗

NLSP: NASA Life Sciences Portal

NASA’s Life Sciences Ports (NLSP) serves the scientific community by providing curated data from space life science experiment. The Human Research Program (HRP) with the help of NLSP is currently transforming their life sciences data archive systems and processes to improve compliance with the FAIR principles. Some of these improvements will at the same time support the twin pillars of Open Science: transparency of methods and reproducibility of results. This video is a high level overview of the NLSP for existing and new users.

Life Sciences data↗

PAHs, Hydrocarbons, and Dimethylsulfides in Asteroid Ryugu Samples A0106 and C0107 and the Orgueil (CI1) Meteorite

Evaluating the molecular distribution of organic compounds in pristine extraterrestrial materials is cornerstone to understanding the abiotic synthesis of organics and allows us to better understand the molecular diversity available during the formation of our solar system and before the origins of life on Earth. In this work we identify multiple organic compounds in solvent extracts of asteroid Ryugu samples A0106 and C0107 and the Orgueil meteorite using two-dimensional gas chromatography and time-of-flight high resolution mass spectrometry (GC×GC-HRMS). Our analyses found similarities between the molecular distribution of organic compounds in Ryugu and the CI carbonaceous chondrite Orgueil. Specifically, several PAHs and organosulfides were found in Ryugu and Orgueil suggesting an interstellar and parent body origin for these compounds. We also evaluated the common relationship between Ryugu, Orgueil, and comets like Wild-2; however, until comprehensive compound-specific isotopic analyses for these organic species are undertaken, and until the effects of parent body processes and Earth’s weathering processes on meteoritic organics are better understood, their parent-daughter relationships will remain unanswered. Finally, the study of organic compounds in Ryugu samples and the curation practices for the future preservation of these unvaluable materials are also of special interest for future sample return missions, including NASA’s OSIRIS-REx asteroid sample return mission.

Hayabusa2↗

Amino Acid Encoding for Deep Learning Applications

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.

Deep-learning↗

Amino Acid Encoding for Deep Learning Applications

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-ofthe-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.

Hesham ElAbd↗

Empowering Open Science with the Science Discovery Engine

: This presentation will describe the work to date in building the SDE as well as what the team has learned about the SMD ecosystem, information curation, and data governance. A short demonstration of the SDE will be presented, and an overview of near- and long-term goals for future development will be shared. Community feedback will be welcomed about the interface, content, and other features to help inform actions to maximize the SDE’s performance and usability. Whether users aim to discover Earth-like atmospheres on planets outside of our solar system or better understand the impacts of solar energy on our own planet, the Science Discovery Engine provides a means for scientists and all curious individuals to find content to further their understanding of science across all time and space scales.

Emily Foshee↗