Bringing the Animal QTLdb and CorrDB into the future: meeting new challenges and providing updated services.
Hu ZL, Park CA, Reecy JM. · 2022
Read Bringing the Animal QTLdb and CorrDB into the future: meeting new challenges and providing updated services. by Hu ZL, Park CA, Reecy JM. (2022) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc8728226.txt.
Introduction
Hu, Park, and Reecy describe updates to the Animal QTLdb and CorrDB, twin resources for livestock genetics. These databases serve the global farm animal research community. They began as a pig-only store for trait-linked gene regions. Years of active growth kept pace with genome science.
The QTLdb now holds 220,401 records linking traits to gene regions. They cover 2,210 traits in seven livestock species. CorrDB holds 23,552 trait correlations and data on 4,273 heritability records. Facts were curated from 3,157 publications. New species, genome builds, and browsing tools were added. Standard mapping steps let users compare data across species.
Catalog record
| CREATOR | Hu ZL, Park CA, Reecy JM. |
|---|---|
| YEAR | 2022 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Bringing the Animal QTLdb and CorrDB into the future: meeting new challenges and providing updated services By Zhi-Liang Hu, Carissa A Park, James M Reecy Published in: Nucleic Acids Research DOI: 10.1093/nar/gkab1116 ABSTRACT Abstract The Animal QTLdb (https://www.animalgenome.org/QTLdb) and CorrDB (https://www.animalgenome.org/CorrDB) are unique resources for livestock animal genetics and genomics research which have been used extensively by the international livestock genome research community. This is largely due to the active development of the databases over the years to keep up with the rapid advancement of genome sciences. The ongoing development has ensured that these databases provide researchers not only with continually updated data but also with new web tools to disseminate the data. Through our continued efforts, the databases have evolved from the original Pig QTLdb for cross-experiment QTL data comparisons to an Animal QTLdb hosting 220 401 QTL, SNP association and eQTL data linking phenotype to genotype for 2210 traits. In addition, there are 23 552 correlations for 866 traits and 4273 heritability data on 1069 traits in CorrDB. All these data were curated from 3157 publications that cover seven livestock species. Along with the continued data curation, new species, additional genome builds, and new functions and features have been built into the databases as well. Standardized procedures to support data mapping on multiple species/genome builds and the ability to browse data based on linked ontology terms are highlights of the recent developments. INTRODUCTION Over the past two decades, livestock genomics research has been greatly accelerated by advances in high-throughput genomics technologies. Not only are whole-genome assemblies available for every major livestock species (1), but multiple genome builds within each species for major livestock breeds have been or are being developed (2). In the meantime, continued discovery of single nucleotide polymorphisms (SNPs) has enabled the generation of high-throughput genotype information from different SNP chip platforms and production of massive genotype–phenotype data from association studies, trait correlation analysis, and heritability estimates (3). This is reflected in the steady data increase in the QTLdb/CorrDB in all seven livestock species (Figure 1). [Figure: Cumulative curated data increase in all seven species hosted at the Animal QTLdb/CorrDB (according to database release records; data transformation was performed on data counts in order for small counts to be visible).] Although efforts to functionally annotate gene expression information are progressing (4), addressing the need for rapid access to and interpretation of all the associated data remains an increasingly challenging task. As the Animal QTLdb has been modified over time to accept curation of multiple data types (e.g. SNP associations, eQTL, copy number variations (CNV), signatures of selection, epistasis, pleiotropy), and curation of correlation and heritability data into the CorrDB has continued (5), the curation workload has increased significantly (Figure 2). We have been able to boost curation efficiency by developing new tools and structures to improve the dataflow with built-in quality controls. The Animal QTLdb and CorrDB have been developed in tandem to minimize duplication of effort. Both databases have undergone three data releases per year since our last report (6). In this paper, we present our ongoing efforts to further develop the QTLdb and CorrDB with a more elastic infrastructure to increase flexibility for inclusion of new types of data entries, new genome builds for genome map data lift-over, new ontology tools for linking information, and additional tools to assist users with viewing and analyzing data. [Figure: A stacked column plot showing a significant increase of curated data by data types in recent years (Notes: (i) the data for 2021 are mid-year counts; (ii) counts for all data types are transformed for small values to be visible; (iii) prior to 2009, the database did not have a full-time curator).] There are five essential data entities in the QTLdb/CorrDB: trait information, map information (QTLdb only), gene information (for eQTL data only), animal information, and publication/experiment information. The major improvements on tools and procedures we have made to facilitate the curation of two of these, trait and map information, will make these databases better positioned to meet upcoming needs. NEW DEVELOPMENTS Database structure improvements and standardized protocol for adding new species to QTLdb/CorrDB The QTLdb was initially developed for a single species (7). Other species were added by using a copy of the database and tool set, then manually adding the necessary parts, such as set of traits, genome/linkage maps, list of markers on these maps, etc., before any data could be curated for that species. To streamline this process, we have developed a standardized protocol for adding a new species which is easily operable as a routine. For this purpose, a set of linked web forms was set up to administer the introduction of a new species, initiate trait entry for the new species (Supplementary data, Figures S1–S3), and introduce a new genome map/assembly (detailed below). The set of web forms serves not only to simplify information input but more importantly to enforce the entry of required information in order to main database integrity, and functions as an information checkpoint (Supplementary data, Figure S4). This pipeline was tested during the recent addition of goat (a new species) and a new genome build for an existing species (sheep) to the QTLdb/CorrDB. Livestock SNP ID/name matching repository to assist SNP-based data curation One of the requirements for entering SNP association data into the Animal QTLdb is that a map location must be represented using a permanent official SNP ID, or ‘rs number’, to ensure data persistence. This means that the SNPs used in experiments/reports must be submitted to, processed, and released by a public SNP archive (non-human organism SNP data was previously hosted at NCBI, https://www.ncbi.nlm.nih.gov/snp/, and is currently at the EMBL-EBI European Variation Archive [EVA], https://www.ebi.ac.uk/eva/). However, due to the exponential increase of SNP data and limited resources to process them, the current release cycle at the EVA can cause very long delays at Animal QTLdb as we wait for the official IDs to become available, creating a bottleneck in our release of curated data. This temporary unavailability of official SNP ‘rs numbers’ interferes with timely data curation and release, because the SNP-based QTL/association experiment is carried out and results published ahead of the EVA release of the SNPs used. Often custom SNP identifiers, rather than universally stable ‘rs’ IDs, are used in QTL/association data publications. Custom SNP IDs/names usually come from SNP discovery workflows, genotyping platforms, and data integration and comparison processes. They may appear in scientific reports prior to SNP data being submitted to or released by a public repository. Not only are there long intervals between submission of SNP data and assignation of stable ‘rs’ IDs, but the original custom SNP identifiers are often not linked to the final SNP IDs. The SNP identity matches became an essential part of our curation of SNP-based QTL/association data. To overcome these difficulties, we reached out to the research community to collect original or secondary data that provide the custom SNP IDs/names along with their ‘rs’ numbers. To date, we have gathered 7 856 530 known SNP ID/name matches in six species and 5 055 768 SNP ‘rs’ to ‘ss’ ID matches in four species, contributed by 10 research groups, labs and individuals (listed here: https://www.animalgenome.org/tools/SNPnmids/). We