Published online 23 September 2005
Published in Crop Sci 45:2269-2272 (2005)
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP BREEDING, GENETICS & CYTOLOGY
A Curated Internet Database of Oat Pedigrees
Nicholas A. Tinker* and
Jitka K. Deyl
Agriculture and Agri-Food Canada, Bldg. 20, C.E. Farm, Ottawa, ON, K1A 0C6, Canada
* Corresponding author (tinkerna{at}agr.gc.ca)
 |
ABSTRACT
|
|---|
Plant pedigree data are valuable for crop improvement and for genetic studies. We describe the development of an extensive, internationally relevant database of oat (Avena sativa L.) pedigrees. The database contains documented records of oat varieties and breeding lines, their ancestors, and various names that have been applied to them. The database is searchable through a simple interface located at http://avena.agr.gc.ca; verified 13 June 2005. This interface allows keyword searches for variety names and creates diagrams showing known ancestors or descendants of any variety.
Abbreviations: COP, Coefficient of Parentage • GMS, Germplasm Management System • ICIS, International Crop Information System
|
INTRODUCTION
|
|---|
MOST CULTIVARS of self-pollinating crop species, including oat, are developed through inbreeding and selection of progeny from a biparental cross. Knowledge of the pedigree (synonymously, the parentage, ancestry or genealogy) of a plant variety (i.e., a cultivar or breeding line) provides several types of valuable information: (i) it can be used to predict potential performance or adaptation; (ii) it can be used to compute coefficients of parentage (COP); (iii) it can be used to plan further crosses for cultivar development; and (iv) it can be used to trace the inheritance of specific gene variants. Baum and Thompson (1970) described benefits of an electronic registry of plant varieties including pedigree charts and provided guidelines for the development of such a registry.
Plant pedigree data have been used frequently to infer the degree of relationship among plant varieties or to trace the inheritance of characteristics. Bickelmann (1989) used pedigree analysis to determine that the occurrence of fatuoids (an undesirable kernel type in oat) is probably unrelated to ancestry. Cowen and Frey (1987) compared genealogical distance among eight oat cultivars with performance of progeny from each cross in a diallel mating design and determined that genealogical distance was a significant predictor of breeding behavior. Souza and Sorrells (1989)( 1991) used pedigree data to construct a matrix of relationships among North American oat cultivars. The above studies provided the basis for comparisons of genealogical distances with genetic distances predicted from genetic marker studies (Moser and Lee, 1994; O'Donoughue et al., 1994).
Although pedigree data are extremely valuable, the routine use of these data is limited by several factors: (i) information may be required from many different sources to trace extended pedigrees; (ii) information from multiple sources must be merged into a single database in anticipation of queries that may require any particular subset of data; (iii) information from different sources may be incomplete, incorrect, conflicting, or ambiguous; (iv) varieties often have multiple names or identifiers; and (v) simple query tools to access pedigree data are often lacking.
Sources of pedigree data for oat are numerous. The most reliable sources are based on formal variety descriptions published in journals. Plant germplasm banks also contain reliable pedigree descriptions for accessions that have been formally submitted by plant breeders. However, pedigrees of many historical varieties and breeding lines must be researched through personal communications or unpublished breeding notes. Secondary sources of pedigree data include data that has been assembled for specific studies, or for older databases. Stanton (1955), Baum (1972), and Coffman (1977) assembled extensive compilations of oat pedigrees. Unfortunately, this material is not available in electronic format. Other oat pedigree databases were assembled for studies described by Souza and Sorrells (1989)( 1991) and Cowen and Frey (1987). This data can be requested from authors in tabular format.
A number of programs and databases have been developed for managing and tracking pedigrees. The International Crop Information System (ICIS), which is used to manage the International Rice Information System (Bruskiewich et al., 2003), is an integrated set of tools for managing information about pedigrees and performance of plant germplasm. Commercial plant breeding software, such as Agrobase (Agronomix Software, Winnipeg, Canada) is also capable of managing and tracing pedigrees. While both types of systems are capable of integrating large amounts of data and they provide complete data management systems for plant breeding, they require substantial commitments of time and expertise.
Other tools may not provide complete data management systems, but they can be used for specific purposes, and they accept file formats that can be created from other data management systems. Peditree (van Berloo, available at: http://www.dpw.wau.nl/pv/pub/peditree/; verified 13 June 2005) provides a powerful interface for viewing pedigree hierarchies and computing COP. It also provides a filter to export data in the format required for Pedigraph (Garbe and Da, 2004). The program KIN (Tinker and Mather, 1993) can be used to compute COP for large data sets when varying assumptions about inbreeding are needed.
Once a significant body of pedigree data has been assembled into a common format, it may be desirable to make the data available to many potential users. Often, this has been done by distributing data files that can be accessed by local software. However, if a pedigree database is to be managed and updated on an ongoing basis, it may be more desirable to create a web-based interface. Although web-based interfaces may have restricted graphics capabilities, there is no requirement for software installation, and they can be accessed easily by a large target audience. Examples of crop pedigrees that can currently be searched to some extent through the internet include potato (Hutten and van Berloo, 2001: http://www.dpw.wau.nl/pv/; verified 13 June 2005), rice (Bruskiewich et al., 2003: http://www.iris.irri.org/; verified 13 June 2005), wheat (http://mendel.lafs.uq.edu.au/; verified 13 June 2005) and triticale (Baum et al., 1990: http://res2.agr.gc.ca/ecorc/tritic/index_e.asp; verified 13 June 2005).
The objective of this study was to develop an accurate, extensive, and internationally relevant database of oat pedigrees and to present the data using a web-based interface that requires little or no training to use.
|
MATERIALS AND METHODS
|
|---|
A pedigree database was developed in Microsoft Access 2000 (Microsoft Corporation, Redmond, WA) using tables similar to those used in the Genotype Management System (GMS) of ICIS. Modifications involved the addition of several fields that were used to track data updates, add comments, and segregate internal data from public records. Data were added and manipulated by queries developed within the MS Access user interface. A web-based interface was developed in the PHP programming language (www.php.net; verified 13 June 2005).
Oat pedigree data were collected from all sources mentioned above. Before adding each data element, the existing database was inspected for records that could cause duplicates, conflicts, or ambiguities. When problems were identified, research was performed to identify the most probable explanation. This involved validation through multiple information sources and/or personal communications with plant breeders. Best efforts were made to distinguish reporting errors from the existence of distinct varieties with identical names, as well as to document the existence of synonyms (including variations in spelling or language) that have been applied to identical varieties. Sources of information were documented for each unique variety, as well as for each name that has been applied to that variety. A single "preferred name" was assigned to each distinct germplasm entry on the basis of, if possible, the registered variety name. Alternate names included translations, spelling variations, abbreviations, and numeric identifiers.
Primary sources of information were obtained and cited when available. Secondary sources of information (e.g., Stanton, 1955, Baum, 1972, Souza and Sorrells, 1988) were used extensively as research tools but were cited as germplasm references only when primary sources were unknown or unobtainable. Secondary sources were cited more frequently as sources of information about alternate names or abbreviations. Comments were added to alert users to conflicting information or other potential interpretations.
|
RESULTS AND DISCUSSION
|
|---|
Information was assembled on germplasm and various names for germplasm from approximately 120 different sources, including personal communications and internal breeding records. Major sources and categories of information are summarized in Table 1. At the time of writing, the database contained over 13000 unique germplasm records, and over 23000 names and identifiers. Since the database is used extensively in the Agriculture and Agri-Food Canada (AAFC) oat breeding programs, approximately half of the germplasm entries included varieties and breeding lines related directly to AAFC oat breeding programs. Most germplasm from other oat breeding programs was incorporated at the level of released cultivars and historical varieties or landraces that appear in pedigrees of released cultivars.
The advantage of using the ICIS data schema is that germplasm records and germplasm names are contained in separate tables, such that each germplasm entity can have an indeterminate number of name records associated with it. In this way, it is easy to re-associate names with germplasm records if germplasm records are merged or separated. Another advantage of this system is that each germplasm entity and each name can have independent authorities. In addition, the use of the ICIS data schema allows compatibility with tools developed by the ICIS project. When queries were written in MS Access or PHP, it was relatively easy to convert data to different formats, such as that used by Peditree.
The web-based user interface was developed so that users could access the information in the database with little or no training and no prior software installation. Both French and English versions of the interface are available and can be toggled with a language button near the top of the page. The interface is located at the following address: http://avena.agr.gc.ca/.
The initial screen presents an option to search for varieties by name. A simple menu option let's users select whether the search term should be matched exactly, anywhere, at the beginning, or at the end of a database entry. By default, a search is performed across all names in the database, but this can also be restricted to preferred names. The search returns a list of names that match the user's query. Each name is linked to a complete passport record for the germplasm to which the name is attached. The passport data and the associated pedigree are shown and described in Fig. 1
.
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 1. Results of query for cultivar ‘Marion (USA)’ in database "Pedigrees of Oat Lines" (http://avena.agr.gc.ca). Passport data includes entry ID (specific to this database), a preferred name, alternate names, species information, cross date, origin, and a reference for the origin. Comments may include anecdotal information about performance, alternate pedigree information and any comments that the database curator feels might help with the interpretation of the data. References for names are documented separately from the reference for the origin, and these are provided as "ref" links beside each name. In this case, the preferred name ‘Marion (USA)’ has been designated by the database curator to distinguish this germplasm from another variety ‘Marion (Canada)’ which is clearly a distinct variety with a different pedigree. By default, each passport is accompanied by a three- or four-generation pedigree. The pedigree is presented in tabular format which can be copied and pasted into a spreadsheet. The final generation of the pedigree is staggered downward to save space in the table. The number of generations can be changed using the selector beside "Ancestry." Alternatively, the reverse pedigree of an entry ("Descendancy") can be listed.
|
|
The simplicity of this database interface has been valuable to engage additional users and contributors of new data. Users have responded favorably to the convenience of having instant access to pedigrees, and many oat breeders have sent detailed breeding records for their own varieties to gain this convenience for additional germplasm. Since germplasm used in variety development is frequently exchanged internationally, the integration of new pedigree data provides excellent opportunities to increase the number of ancestors that can be traced and thereby increase the accuracy of estimates of COP among varieties. Many modern oat cultivars can be traced back more than 10 generations. The addition of new data can also increase the number of errors contained in a database; therefore, this type of data requires intensive curation. In some cases, errors or discrepancies may not be detectable. For example, two historical pedigrees may list the same ancestral landrace, but the name may have been given arbitrarily to two different sources of seed. Or, a common ancestor may be listed under different names in different pedigrees. Often, however, this is due to language-related translations (e.g., ‘Golden Rain’, ‘Goldregn’, ‘Pluie d’Or') and can be sorted out by finding multiple sources of the same pedigree.
Ongoing maintenance and improvement of this database is planned, and additional contributions or corrections are welcomed.
|
ACKNOWLEDGMENTS
|
|---|
Generous financial support and encouragement for this project was provided by the Quaker Oats Company and Quaker Gatorade Tropicana Canada Inc. We are grateful to many researchers who spent significant amounts of time responding to enquiries related to the database, and to those who sent new information to be included. We thank Drs. A.R. McElroy and S. Molnar for project guidance, Anissa Lybaert and Philippe Couroux for translations, and Christopher Ross for contributions to the database interface. Mention of commercial software in this publication does not constitute an endorsement by the authors, their institution, or the publishers of this journal. ECORC-AAFC Contribution Number 04-457.
Received for publication November 30, 2004.
|
REFERENCES
|
|---|
- Baum, B.R., and B.K. Thompson. 1970. Registers with pedigrees for cultivars; their importance, their contents, and their preparation by computer. Taxon 19:762–768.[CrossRef]
- Baum, B.R. 1972. Material for an international oat register. AAFC Publication A52–4772. Agriculture and Agri-food Canada, Ottawa CA.
- Baum, B.R., L.G. Bailey, J. Nugent, W.G. Royds, and R. Elvidge. 1990. The Triticale cultivars international register computer system, a flexible information system suitable for international registration. Taxon 39:9–15.[CrossRef]
- Bickelmann, U. 1989. The pedigree of German oat cultivars (Avena sativa L.) and the occurrence of fatuoids. Plant Breed. 103:163–170.[CrossRef]
- Bruskiewich, R.M., A.B. Cosico, W. Eusebio, A.M. Portugal, L.M. Ramos, M.T. Reyes, M.A. Sallan, V.J. Ulat, X. Wang, K.L. McNally, H.R. Sackville, and C.G. McLaren. Bioinformatics. 2003. Linking genotype to phenotype. The International Rice Information System 19(Suppl. 1):i63–i65 (IRIS).
- Coffman, F.A. 1977. Oat History, Identification and Classification. USDA-ARS Tech. Bull. 1516. U.S. Gov. Print. Office, Washington, DC.
- Cowen, N.M., and K.J. Frey. 1987. Relationship between Genealogical Distance and Breeding Behaviour in Oats Avena-Sativa L. Euphytica 36:413–424.[CrossRef]
- Garbe, J.R., and Y. Da. 2004. Pedigraph 2.0, A software tool for the graphing and analysis of large complex pedigrees. p. 242. In Abstract book. ADSA-ASAS-PSA Joint Annual Meeting, St. Louis. 25–29 July 2004.Hutten, R.C.B., and R. van Berloo. 2001. An online potato pedigree database. URL: http://www.dpw.wau.nl/pv/newquery.asp; verified 29 June 2005.
- Moser, A.U., and M. Lee. 1994. RFLP variation and genealogical distance, multivariate distance, heterosis, and genetic variance in oats. Theor. Appl. Genet. 87:947–956.
- O'Donoughue, L.S., E. Souza, S.D. Tanksley, and M.E. Sorrells. 1994. Relationships among North American oat cultivars based on restriction fragment length polymorphisms. Crop Sci. 34:1251–1258.[Abstract/Free Full Text]
- Souza, E., and M.E. Sorrells. 1988: Coefficients of parentage for North American oat cultivars released from 1951 to 1986. New York Agric. Exp. Sta. No. 33, Cornell University, Ithaca, NY.
- Souza, E., and M.E. Sorrells. 1989. Pedigree analysis of North American oat cultivars released from 1951 to 1985. Crop Sci. 29:595–601.[Abstract/Free Full Text]
- Souza, E., and M.E. Sorrells. 1991. Relationships among 70 North American oat germplasms. I. Cluster analysis using quantitative characters. Crop Sci. 31:599–605.[Abstract/Free Full Text]
- Stanton, T.R. 1955. Oat Identification and Classification. USDA Technical Bulletin 1100. U.S. Gov. Print. Office, Washington, DC.
- Tinker, N.A., and D.E. Mather. 1993. KIN: Software for computing kinship coefficients. J. Hered. 84:237.[Free Full Text]
TOP
ABSTRACT
INTRODUCTION
