Computer-supported gene bank management

Summary

Man's interest in plants goes back to times immemorable. The collecting of useful plants to improve the output from agriculture is likewise very ancient. A clear purpose to utilize became apparent at the end of the last century and a purpose to also preserve came withVavilov in the 1930-ies.

Preservation became a pressing necessity as the agricultural development after the Second World War accelerated and this necessity became very apparent as the Mexican wheats were introduced, which resulted in systematic efforts to establish national and regional gene banks all over the world.

It soon became clear that the establishment of gene banks among other things led to a massive accumulation of data which could not possibly be handled with traditional manual field book and recording systems. The computer was introduced to support the information handling.

This lecture presents an overview of the major components to consider in designing a computerized gene bank management system. One consideration forwarded is that such a system should be utilizable not only on a recording and retrieval-level (R&R-level) but also used for analysis, producing new data for further interpretation (A-level).

To ensure utility as well as safety aspects, a collection of genetic resources material could be preserved as different collections-alternatively as parts of one collection-for different purposes, Duplicate Base Collection for long-term safety, Base Collection for rejuvenation, Active Collection for multiplication and distribution, characterization, research, etc., the organization implemented to be reflected in the designing of databases.

An attempt is made to discuss what material should come under the responsibility of gene banks for preservation. Arguments are given for excluding plant breeding material, plant introductions, commercial seed and ordinary agricultural products from the gene bank's responsibilities.

It is further argued that the samples, accessions, preserved in the gene bank are vehicles for the genes, which are the units of variation, recombined and selected on. The justification for preserving an accession can therefore be expressed in terms of it being a vehicle for specific genes, geneblocks, chromosomes, genotypes or groups of genotypes.

The preservation of genetic resources may be handledex situ orin situ, and an accession can be preserved in either way or by a combination of both, the most appropriate mode depending on what the accession is supposed to preserve, which again has to be considered in designing a management system.

The information relating to the material to be stored and handled by a gene bank management system is often categorized, based on content and use, into passport, management and character data. An argument is put forward in favour of regarding gene and taxonomic data also as categories of their own, since the utilization of such data can become a powerful tool but will require specifically programmed software.

The description of an accession in terms of genes using gene symbol descriptors is considered the superior method, to which taxonomic infraspecific classification offers a good substitute for species lacking detailed genetic information. Least preferable is to rely completely on unorganized character descriptors; it is pointed out, however, that carefully designed character descriptors in combination with competent observations for those descriptors would result in information needed to elaborate an infraspecific classification.

Finally, some more important auxiliary databases used in a computerized gene bank management system are briefly mentioned, for instance code keys in general and such for genes and taxa in particular.

Man's interest in plants in general and cultivated plants in particular goes back to times immemorable, for obvious reasons. Collecting of plants for purposes that we would to-day call plant breeding has probably also taken place for thousands of years though records of these activities are scarce.

At the end of the 19th century, after the work of theVilmorin-brothers became known and plant breeding activities were attempted in different European countries, systematic and sometimes well recorded collecting activities were initiated. Land-races of most crops were collected and screened, selected upon and crossed with. After the rediscovery ofMendel's results the breeding work, having obtained a scientific basis, accelerated and in the two first decades of this century new cultivars appeared to replace, at an increasing rate, the old land-races. In this process, little attention was given to the material brought together-modern material was to replace the old and obsolete, and what became of the old and obsolete was of no great concern.

A change in this attitude came in the 1930-ies withVavilov and his scientific contributions. Among other things he made clear that the old land-races were, in fact, a phylogenetic record to be interpreted and a valuable, adapted material for use in plant breeding. Systematic collecting activities to cover different geographic areas were again carried out but this time with the clear intention to save the material for the future; the collections then established are the first gene banks, though that name was not yet thought of.

The gene bank concept was borne as a consequence of the realization that the post-war efforts to increase the per-capita agricultural production in the world had to include a massive input in plant breeding resulting in new varieties which would threat with extinction a large part of the hitherto cultivated material. The more general awareness was, however, slow in coming and several examples of emergency situations were encountered, for instance when the new Mexican wheats began to spread. It was only in 1974 that the Consultative Croup on International Agricultural Research (CGIAR) sponsored by FAO, the World Bank and United Nations Development Programme established the International Board for Plant Genetic Resources (IBPGR) to aid in the world-wide preservation of plant genetic resources. One of the means used by IBPGR became the support for establishing national and regional gene banks.

Many gene banks with a history as what may be called Vavilovian seed collections had well established management systems, usually manual, while many newly established gene banks had no tradition, no system at all to fall back on, neither manual, nor computerized. In fact, a generally agreed, scientifically founded genetic resources concept, giving genetic resources preservation a scientific standing similar to e.g. ecology, has yet to be presented and implemented.

This explains largely the present situation with respect to computer supported gene bank management systems. Despite efforts from different organizations to uniformize and coordinate the work, almost every gene bank that acquired computers and acquired or developed software started out by computerizing whatever manual system that was already in use or, where no system existed, were satisfied with piecemeal solutions to the most pressing problems, offered by available standard software packages.

Within the framework of consultancies to IBPGR and to the Nordic Gene Bank and as a plant breeder having also since 1962 served as curator for a quite old collection with a since 1973 completely computerized management system, the author has had opportunity to give the problem considerable attention. The following presentation therefore contains elements of a gene bank management system which have been tested and implemented as well as such which have not yet but are considered essential for a system adequately meeting scientific as well as practical needs.

Firstly must be emphasized that bringing in the computer as a tool in the gene bank management introduces an enormous new potential in processing data and handling information, not only quantitatively but also qualitatively. The computer can, of course, do nothing that can not be done manually and by using the brain, since the computer works with a software produced by man (leaving out for the time being the more extreme prophecies on Artificial Intelligence).

In contemplating how to design a computerized gene bank management system it is very useful to think in terms of computerization on different levels:

• -The recording and retrieval level (R&R), at which the data and the processing are not used for creating new data, and

• -the analytical level (A), where data are used to create new data and new information to be interpreted.

Most of the discussion going on to-day regarding computerization of gene bank management relates to the R&R-level at which advantages and gains are mainly of a practical nature, such as timesaving and better overview and insight. Most of the scientific advantages of the computerization are, however, to be had at the A-level. This, however, requires that the practical management of the material is so organized as to yield high level quality data with high level information content. Needless to say, costs and competence required are major factors to be considered. Competence is expensive, and particularly so in the field of computer software and this in itself can be enough to justify for the gene banks to develop inhouse computer competence. The decisive factor to strongly recommend the gene banks to have an own programming competence is really another, viz., that programming for the A-level and for an A-level prepared R&R-level requires a combination of biological specialized gene bank competence combined with the computer competence, since the A-level calls for biologically “intelligent” programs.

The first considerations concern the organization of collections. In the long term, one has to achieve adequate longevity of the stored material, safety from losses by accidents, adequate quantities for distribution, efficiency in recording data and economy with funds and competence. At the same time the current situation needs to be considered, since changes in an existing gene bank management system necessarily have to be evolutionary rather than revolutionary. The following kinds of collections—or functions within one physical collection—are considered normally to be needed in a full-scale gene bank.

Zusammenfassung

Die Entwicklung der Menschheit ist von Anbeginn mit den Pflanzen verbunden. Das Sammeln nutzbarer Pflanzen, um die Erzeugung aus der Landwirtschaft zu erhöhen, ist ebenfalls sehr alt. Am Ende des letzten Jahrhunderts begann man, Pflanzen für die planmäßige Nutzung in der Züchtung zu sammeln.Vavilov ging in den 30er Jahren darüber hinaus und betonte, daß die gesamte Mannigfaltigkeit kultivierter Pflanzen zu sammeln und zu erhalten ist.

Dringend erforderlich wurde ihre Erhaltung, als sich nach dem Zweiten Weltkrieg die Landwirtschaft beschleunigt entwickelte und besonders mit der weltweiten Verbreitung der mexikanischen Weizen das Wirken der Generosion offenbar wurde. Das führte schließlich zu Bestrebungen, planmäßig überall in der Welt nationale und regionale Genbanken einzurichten.

Bald wurde man gewahr, daß die Einrichtung von Genbanken, neben anderen, zu einer massiven Anhäufung von Daten führte. Sie konnten unmöglich mit den traditionellen, manuell geführten Feldbüchern und Dokumentations-systemen gehandhabt werden. So wurde der Computer eingeführt, um die Bearbeitung der Informationen zu unterstützen.

Dieser Vortrag gibt einen Überblick über die wichtigeren Komponenten, die bei der Planung eines rechnergestützten Systems für das Genbank-Management zu berücksichtigen sind. Eine Überlegung bezieht sich darauf, daß ein derartiges System nicht nur zur Aufzeichnung und Wiedergabe von Informationen (recording and retrieval, R&R-level) verwendet werden sollte; es sollte auch für Analysen und die Erzeugung neuer Daten zu nutzen sein, aus denen sich weitere Interpretationen ableiten lassen (A-level).

Das Material genetischer Ressourcen kann sowohl unter Berücksichtigung von Gesichtspunkten seiner Nutzung als auch der Sicherheit seiner Erhaltung in unterschiedlichen Kollektionen — alternativ als Teile einer Kollektion — erhalten werden: Basis-Kollektion von Duplikaten für eine langfristige sichere Erhaltung, Basis-Kollektion für die Erneuerung des Materials, Aktive Kollektion für die Vermehrung und Verteilung, für die Charakterisierung, für die Untersuchung des Materials usf. Die damit verbundene praktische Organisation der Kollektion muß sich in der Planung der dazugehörigen Datenbank widerspiegeln.

Es wird versucht darzulegen, für welches Material Genbanken die Verantwortung zur Erhaltung übernehmen sollten. Gründe für den Ausschluß von der Erhaltung in Genbanken werden für das Zuchtmaterial, für introduzierte Pflanzen, für Material von Handelssorten und aus der normalen landwirtschaftlichen Produktion dargelegt.

Weiterhin wird erörtert, daß die in der Genbank erhaltenen Muster Träger von Genen sind. Die Gene sind, rekombiniert und selektiert, die Einheiten der Variation. Die Rechtfertigung für die Erhaltung eines Musters läßt sich daher so ausdrücken, daß es ein Träger für spezifische Gene, Genblöcke, Chromosomen, Genotypen oder Gruppen von Genotypen ist.

Es ist möglich, die genetischen Ressourcenex situ oderin situ zu erhalten. Ein Muster kann entweder auf eine der beiden oder in einer Kombination beider Arten erhalten werden. Das geeignetste Verfahren hängt von dem Grund der Erhaltung des betreffenden Musters ab. Die Art und Weise der Erhaltung muß bei der Planung eines Management-Systems berücksichtigt werden.

Die auf das Material bezogene Information, die mit einem Genbank-Management-System gespeichert und bearbeitet werden muß, wird oft in Kategorien eingeteilt. Dem Inhalt und der Verwendung entsprechend ordnet man sie den Passport-, Management- und Charakterisierungsdaten zu. Die Einführung eigener Kategorien für die genetischen und taxonomischen Daten wird begründet. Ihre Nutzung kann sich als sehr wertvoll erweisen; sie erfordert aber eine besonders programmierte Software.

Die Beschreibung eines Musters durch die Angabe von Genen, unter Verwendung von Gensymbol-Deskriptoren, wird als die weiter fortgeschrittene Methode angesehen. Dazu bietet für die Arten, bei denen eine detaillierte genetische Information fehlt, die infraspezifische Klassifikation einen guten Ersatz. Am wenigsten ist zu empfehlen, sich vollständig auf ungeordnete Merkmalsdeskriptoren zu stützen. Es wird jedoch darauf verwiesen, daß sorgfältig definierte Merkmalsdeskriptoren in Verbindung mit angemessenen Beobachtungen die für die Ausarbeitung einer infraspezifischen Klassifikation benötigten Informationen ergeben würden.

Am Ende werden einige wichtige Hilfsdatenbanken kurz erwähnt, die in einem rechnergestützten System für das Genbank-Management gebraucht werden. Dazu gehören beispielsweise allgemeine und spezielle Code-Schlüssel, besonders solche für Gene und Taxa.

Краткое содержание

Сбор иолезных растений, с целью повышения продуктнвности селъского хозяйства, тесно связан с нсторией развития человечества. В конце иоследнего столетия и, особенно, после второй мнровой войны, в связи с ускоренным развитием сельского хозяйства, влекущим за собой генную зрозню, стала очевидно, указываемая ещё в 30-х годах Н. И. Вавиловым, необходимость сбора и сохранення всеобщего разнообразия растительного мира. С этой целью ио всему миру были создаиы национальные и региональные генные банки.

Однако, создание генных банков вело к накоилению большого массива данных, которые было бы невозможно обрабатывать по традиционным документационным системам, с ведущимися от руки иолевыми книгами. Отсюда следующий шаг — иривлечение ЭВМ для обработки информации.

В данном докладе даεтся обзор наиболее важных компонент, которые следует учитывать ири развитии собственной программной системы для управления генным банком. Прежде всего, следует учесть, что подобные системы должны служить не только для регистрации н воспроизведения информации (recording an d retrieval, R&R-level), но и для εё анализа, получения новых данных с их (иоследующей интерпретацией (A-level).

С учетомаспектов использования и надежности, матернал генетических ресурсов можно хранить в коллекциях различных тинов — нли же в различных частях одной коллекции —, прелназначенных для различных целей: в базисной коллекции дубликатов для длнтельного обеспечения сохранности, в базизсной коллекции для восироизведения материала, в рабочей коллекции для размноюения и обмена, для иеследовательских целей, и т. д. Практическая организация коллекции, отвечающая конкретным заданиям, должна найти свое отражение в организации сответствующего банка данных.

Деластся нопытка объяснить какой внд матернала нодлежнт хранению в генных банках. Так как образец, хранимый в генном банке, является носнтелем генетнческой информации, то обоснованием для его хранения межет служить следующее утвержденне: образец — это неснтель снецифнческнх генов, генных блоков, хромосом, генотнна нлн грунны генотннов.

Образцы генетнческнх ресурсов можно храннть ех situ нлн in situ, либо обонмн снособами, но тии хранення следует учитывать ири органнзацнн нрограммной снстемы уиравлення.

Ннформация, касающаяся матсрнала, храннмого в намятн автсматнзнрованного банка данных, нодразделяется на категорнн. Согласно содержанню н нснользованню выделяют наснортные, менеджементные н характернстнчеекне данные. Обоеновывается введение собственных категорий для генетнческнх н таксономическнх данных. Однако, их нснользованне требуст есебсго матсматнческого обеснечення.

Оннсанне образца нутем ввода генетнческой ннформацнн, нснользуя для этого дескрииторы генных снмволов, рассматрнвается как более нрогресснвный метод. Для образцов, у которых отсутствует нодробная генетнчсская ннформация, хорошей заменой служнт внутрнвндовая класснфнкация.

В заключение уномннаются некоторые важные всномогательные банки данных, которые необходнмы нрн работе с геннымн банкамн. К ним, нанрнмер, относятся общне н снецнальные кодовые ключн, в частности, для генсв и таксонов.