Remember the big discussion about how to ensure that information about genebank accessions can be linked back to the accession itself? Our friend Dr Dag Endresen has written a handy guide to Digital Object Identifiers (DOIs) as they might be used by genebanks. In addition to explaining how the system works, he points out that it might be a good idea for one organization to fork over the roughly USD1000 to register a top-level DOI name such as genesys. (Bioversity? The Trust? The Treaty? Is anyone listening?) That would not stop any other genebank flush with cash from registering their own domain, and it would provide something to the data providers at genebanks in return for their data. Dag discusses some other options on his blog, and I’m sure there’ll be lots of discussion there, or here, or some other place. Either way, the sooner some similar system is adopted, the sooner we can trace our collective way out of genebank database hell and satisfy the needs of those who want to link data to accessions.
2 Replies to “Genebank data identifiers”
Thanks for your as always excellent comments! The text I blogged comes in part from a project meeting at NordGen last week and in part from a discussion with Michael and Theo last November – and started with what to put on barcodes to be printed on seed bags. I was also thinking of the discussion here on germplasm documentation as a two-way street as you also mention. One big problem for the utilization of genebank collections is the lack of an effective link between the genebank accessions and descriptive data such as trait observations – often generated by other stakeholders than the genebanks who are keeping the germplasm collections. A mechanism like DOIs (or other persistent identifiers) could be the key technology to make dynamic aggregation of passport data and descriptive data possible. The second social obstacle, if people *want* to share their descriptive information is another issue ;-)
This discussion is of interest to me as someone maintaining a collection of DNA samples with no living material associated. The original living collection of taro that I worked with in 1985-1990 has long since been abandoned, but I still keep the DNA samples and a certain amount of accession passport data. I have recently unified my list of samples from scattered notes, and wonder whether I should add DOI numbers. For practical purposes, the collection may have continue to have more permanence than living collections because it is possible to send out highly diluted sub-samples for various DNA analyses while maintaining the original stocks in the collection. This is an unexpected benefit of the PCR revolution in DNA analysis. A 1 ml sample of DNA at 1ug/ul can live for ever, almost, when full genome sequences can be produced using just a tiny fraction of the DNA contained in the tube. In my project on wild taro in Asia, I am encouraging colleagues to develop and keep their own DNA archives, to reduce the need for procedures for international exchange of genetic materials. Each owner of a collection can eventually request analysis of samples inside or outside their own country, a process that does not involve giving away control of the permanent stock samples. However, this kind of de-centralised approach to archive management and collaborative research will require good record keeping, and DOIs may be the answer.