The variable quality of metadata about biological samples used in biomedical experiments

Gonçalves, Rafael S.; Musen, Mark A.

doi:10.1038/sdata.2019.21

Cited by 75 publications

(77 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This problem is also common to other fields of research. A recent metaanalysis of 11.4 million metadata of samples used in biomedical experiments revealed multiple anomalies and highlighted the need for a more robust approach to reporting metadata [36]. Inclusion of metadata is crucial for finding suitable data sets for meta-analyses, will help to ensure interoperability of the metadata, and ultimately will allow us to begin to understand how the host plant (the environment of the rhizosphere) shapes macroecological patterns in rhizosphere microbiomes.…”

Section: Can We Transcend Taxonomic and Geographic Limitations?mentioning

confidence: 99%

Towards Unraveling Macroecological Patterns in Rhizosphere Microbiomes

Brunel

Pouteau

Dawson

et al. 2020

Trends in Plant Science

View full text Add to dashboard Cite

It is generally accepted that plants locally influence the composition and activity of their rhizosphere microbiome, and that rhizosphere community assembly further involves a hierarchy of constraints with varying strengths across spatial and temporal scales. However, our knowledge of rhizosphere microbiomes is largely based on single-location and time-point studies. Consequently, it remains difficult to predict patterns at large landscape scales, and we lack a clear understanding of how the rhizosphere microbiome forms and is maintained by drivers beyond the influence of the plant. By synthesizing recent literature and collating data on rhizosphere microbiomes, we point out the opportunities and challenges offered by advances in molecular biology, bioinformatics, and data availability. Specifically, we highlight the use of exact sequence variants, coupled with existing and newly generated data to decipher the rules of rhizosphere community assembly across large spatial and taxonomic scales. Unearthing the Macroecology of Rhizosphere Soil Microbes Recent advances in sequencing technologies have expanded our ability to study plantassociated microbial communities. This has transformed our perception of the interactions between the plant and its microbiome (see Glossary) [1], which are now increasingly regarded jointly as holobionts [2-5]. The rhizosphere (i.e., the interface of plant roots and soils [6]) hosts diverse communities of microorganisms that are crucial to the plants they associate with. The rhizosphere microbiome can supply plants with nutrients [7], and protect plants against pathogens [8]. Furthermore, microbiomes can stimulate plant growth by producing phytohormones [9,10], and improve plant resistance and tolerance to abiotic stressors. Recent research suggests that rhizosphere microbiomes can alter plant phenology (e.g., flowering time) [11], modify morphological and size-related traits (e.g., shoot and root length and biomass, and number of secondary roots and leaves) [12], have a major role in plant community dynamics [13-15], and mediate plant responses to global change [16,17]. While much progress has been made, we are still far from understanding the mechanisms that control rhizosphere microbiome assembly and maintain community structure and composition. It has long been hypothesized that the macroecological patterns (i.e., ecological patterns across large spatial scales) of the rhizosphere microbiome will relate to the macroecological patterns of plants [18]. However, few studies have found clear relationships between plant and rhizosphere diversity [19,20]. This is likely because there is a disconnect between the plant and the microbial scales (Box 1). For other (non-rhizosphere) microbiomes, large-scale sampling campaigns are leading the way by generating standardized raw data and metadata (e.g., the Earth Microbiome Project [21] or the Human Microbiome Project [22]). For macro-organisms, macroecological patterns are frequently identified by collating existing and newly generated data from ...

show abstract

Section: Can We Transcend Taxonomic and Geographic Limitations?mentioning

confidence: 99%

Towards Unraveling Macroecological Patterns in Rhizosphere Microbiomes

Brunel

Pouteau

Dawson

et al. 2020

Trends in Plant Science

View full text Add to dashboard Cite

show abstract

“…User-defined fields, though useful and even necessary in certain situations, have led to a significant increase in heterogeneity across this dataset and others (5). The use of word embeddings for clustering attributes by semantic similarity revealed a lack of normalization in attribute naming, mostly in the form of small deviations in spelling and punctuation (e.g., cell type and Cell type).…”

Section: Discussionmentioning

confidence: 99%

“…(5) recently described the variable state of the metadata available in databases such as NCBI's BioSample and the European Bioinformatics Institute's (EBI) BioSamples (6). The infrequent use of controlled vocabularies in the metadata submission process, coupled by the allowance for the creation of user-defined attributes, has led to an explosion of heterogeneity in the overall metadata landscape (5). This can often hinder researchers' ability to fully utilize the potential information that a given dataset, or a meta-analysis of multiple datasets, might hold.…”

Section: Introductionmentioning

confidence: 99%

Increasing metadata coverage of SRA BioSample entries using deep learning based Named Entity Recognition

Klie

Tsui

Mollah

et al. 2018

Preprint

View full text Add to dashboard Cite

Motivation:Extraction of biomedical knowledge from unstructured text poses a great challenge in the biomedical field. Named entity recognition (NER) promises to improve information extraction and retrieval. However, existing approaches require manual annotation of large training text corpora, which is laborious and time-consuming. To address this problem we adopted deep learning technique that repurposes the 43,900,000 Entity-free-text pairs available in metadata associated with the NCBI BioSample archive to train a scalable NER model. This NER model can assist in biospecimen metadata annotation by extracting named-entities from user-supplied free-text descriptions.Results: We evaluated our model against two validation sets, namely data sets consisting of short-phrases and long sentences. We achieved an accuracy of 93.29% and 93.40% in the short-phrase validation set and long sentence validation set respectively.Availability: All the analyses, pre-trained model, environments, and Jupyter notebooks pertaining to this manuscript are available on Github: https://github.com/brianyiktaktsui/DEEP_NLP . Contact: hkcarter@ucsd.eduFig1. Repurposing public biospecimen data for NER training ( A ) Depiction of training NER model using pre-annotated Entity-free-text pairs available from public biospecimen annotation data (BioSamples) from NCBI ( A.1 ) Example of Entity-free-text pairs from BioSamples. In this example, the free-text phrase Glioblastoma stage 4 system is a Disease entity. ( A.2 ) Expected results of an NER model recognizing biomedical concepts from sentences. ( B ) Histogram of the 30 most frequently used entities (x-axis) available in the current set of BioSamples. These atomic named entities (blue labels) can be used to extract concepts from composite entities TITLE and DESCRIPTION (red labels).

show abstract

“…We constructed an evaluation pipeline to drive the analysis workflow ( Figure 8). This pipeline consists of 7 sequential steps: (1) content download from NCBI BioSample and EBI BioSamples databases; (2) template design for each of those databases and generation of the corresponding template instances; (3) linkage of template instance field names and values to ontology terms; (4) dataset splitting into training and test sets; (5) generation of rules to drive the recommendations from the training set; (6) accuracy measurement using the test set; and (7) results analysis. These steps are now described in more detail.…”

Section: Discussionmentioning

confidence: 99%

Using association rule mining and ontologies to generate metadata recommendations from multiple biomedical databases

et al. 2019

View full text Add to dashboard Cite

Metadata—the machine-readable descriptions of the data—are increasingly seen as crucial for describing the vast array of biomedical datasets that are currently being deposited in public repositories. While most public repositories have firm requirements that metadata must accompany submitted datasets, the quality of those metadata is generally very poor. A key problem is that the typical metadata acquisition process is onerous and time consuming, with little interactive guidance or assistance provided to users. Secondary problems include the lack of validation and sparse use of standardized terms or ontologies when authoring metadata. There is a pressing need for improvements to the metadata acquisition process that will help users to enter metadata quickly and accurately. In this paper, we outline a recommendation system for metadata that aims to address this challenge. Our approach uses association rule mining to uncover hidden associations among metadata values and to represent them in the form of association rules. These rules are then used to present users with real-time recommendations when authoring metadata. The novelties of our method are that it is able to combine analyses of metadata from multiple repositories when generating recommendations and can enhance those recommendations by aligning them with ontology terms. We implemented our approach as a service integrated into the CEDAR Workbench metadata authoring platform, and evaluated it using metadata from two public biomedical repositories: US-based National Center for Biotechnology Information BioSample and European Bioinformatics Institute BioSamples. The results show that our approach is able to use analyses of previously entered metadata coupled with ontology-based mappings to present users with accurate recommendations when authoring metadata.

show abstract

The variable quality of metadata about biological samples used in biomedical experiments

Cited by 75 publications

References 23 publications

Towards Unraveling Macroecological Patterns in Rhizosphere Microbiomes

Towards Unraveling Macroecological Patterns in Rhizosphere Microbiomes

Increasing metadata coverage of SRA BioSample entries using deep learning based Named Entity Recognition

Using association rule mining and ontologies to generate metadata recommendations from multiple biomedical databases

Contact Info

Product

Resources

About