A tribute to the late Professor J.H. Edwards, the creator of Oxford Grids The concept of an Oxford Grid was developed by John Edwards during his tenure as Professor of Genetics at the University of Oxford (1979-1995). Despite his considerable mathematical skills, John was a very keen advocate of the power of simple diagrams to convey important ideas. One of his major contributions to science was to conceive, in the early 1980s, a diagram that has become an exceedingly powerful tool for comparative mapping. The diagram comprises a square divided into rows and columns, with each row representing a chromosome in one species, and each column representing a chromosome in a second species. Chromosomes are presented in numerical order from top to bottom (for the first species) and from left to right (for the second species), with the relative height of each row and the relative width of each column indicating the relative size of that chromosome. The rows and columns divide the grid into rectangles, each representing a combination of two chromosomes (one from each of two species), with the length of the sides of each rectangle proportional to the length of those two chromosomes. In the early days, when orthologous loci were mapped only to whole chromosomes, each orthologue was indicated by a point arbitrarily or randomly placed within the relevant rectangle to display the general pattern of homology. As positional data became available, it became possible to indicate location within each of the two chromosomes. The first such diagram appeared in the second of a series of papers on mouse-human orthologies1 from the Genetics Laboratory in Oxford and the nearby MRC unit at Harwell. This diagram was reproduced by V.A. McKusick in the 8th edition of Mendelian Inheritance in Man (p. cxvi)2, where it was described (for the first time) as an Oxford grid, acknowledging the laboratory in which the concept had arisen. From the very beginning, John had the vision of harnessing the power and utility of computers to (a) store lists of orthologues for any number of pair-wise species combinations, and (b) draw grids automatically3. In the first half of the 1990s, under John’s leadership and in collaboration with Jean Thierry-Mieg (Montpelier, later Bethesda), the powerful C. elegans freeware database ACeDB was adapted (primarily by Jo Dicks) to bring John’s vision to reality, and to extend the vision to include pairwise chromosome grids (an expanded single rectangle of the grid), one-to-many grids (one chromosome of one species versus all relevant chromosomes of the other species), and multi-species grids (one chromosome of one species versus all other species)4,5 (http://ukcrop.net/software/displays.html). Extension to the farmyard John’s development of the concept of Oxford grids was a reflection of his long-standing interest in comparative genomics as a means of gaining knowledge about human inherited disorders. Whilst his initial comparative interests were stimulated in part by his collaboration with Tony Searle, Mary Lyon and other mouse geneticists at Harwell (see, e.g. ref 6), he could see the potential of comparative studies beyond just mice and humans. As he stated in such typical style in a 1994 review, “Unfortunately, this ‘homunculus’ [i.e. the mouse], although convenient to house and cheap to feed, is too small and short-lived to be suited to the study of some human disorders. Venipuncture is difficult, organ transplantation almost impossible, lung biopsy fatal and the skin concealed by hair. This is no criticism of the mouse, which, like ourselves, has virtues and limitations; fortunately limitations in one species are largely balanced by advantages in the other. Sheep, goats, cows and pigs are of a convenient size for therapeutic trials and 100 million of each are available and under scrutiny, their health often supervized by various inspections and their diseases well documented.”7 Consequently, by the time John retired from Oxford in 1995, he was hot on the trail of what he delighted in calling “the farmyard”. At every opportunity anywhere in the world, John worked towards an expanded vision, shared with several others, of a single web resource that would combine the best possible map for each species, available freely on the web in a format facilitating comparisons of maps and loci and traits among species, with each locus/trait hyperlinked to each of the catalogues of genes and traits in particular species. By a fortunate coincidence, by the 1990s two of his children happened to be living in the antipodes — a part of the world with an abundance of farmyard animals. After several stints in New Zealand working with AgResearch colleagues such as Tom Broad, in 1993 John began the first of what were to become mostly annual visits to the Faculty of Veterinary Science at the University of Sydney, where he was appointed Visiting Professor. For the next 13 years, until only months before his death, he collaborated with Frank Nicholas and colleagues mentioned below in creating a web-based resource that brought his expanded vision to fruition: the Oxford Grid Project, hosted by the Australian National Genomic Information Service (ANGIS) at http://oxgrid.angis.org.au/. The programming for the web site and web tools was performed by Stefan Gregory during his time at ANGIS. More recently, Matthew Hobbs (CRC for Innovative Dairy Products) and Jonathan Usmar (SheepGenomics) have further developed the site and the tools. In addition to being on the farmyard trail, for many years John was also hot on the trail of researchers who do not make their maps publicly available in a useful format: to John’s ever-increasing frustration, it seemed that as linkage and physical maps across the mammalian kingdom (and beyond) became better characterised and richer with information, so too they became less accessible to other researchers, typically being available only as poor-quality static images. From its early days of development at Southampton University, John was a keen champion of the Location DataBase (LDB) strategy8,9, because it alone was based on the principle of providing free public access to maps comprising tables of actual locations. The LDB strategy also had the virtue of providing a rational basis for creating a single integrated tabular map for each species, this being an essential pre-requisite for the most informative comparative maps depicted as Oxford grids. Fortunately, John lived to see the LDB strategy expanded into the farmyard10, and for this strategy to be used to enhance the power of comparative mapping via the creation of grids within the Oxford Grid Project. Frank Nicholas 11 December 2007 1. Buckle, V.J., Edwards, J.H., Evans, E.P., Jonasson, J.A., Lyon, M.F., Peters, J., Searle, A.G. & Wedd, N.S. Chromosome maps of man and mouse II. Clin. Genet. 26, 1-11 (1984). 2. McKusick, V.A. Mendelian Inheritance in Man. Catalogs of Human Genes and Genetic Disorders. (8th edition) Baltimore: Johns Hopkins University Press (1988). 3. Edwards, J.H. The Oxford grid. Annal. Hum. Genet. 55: 17-31 (1991). 4. Dicks, J. The display of comparative mapping data using ACEDB. Genetic Mapping of Disease Genes, Pawlowitski, I-H., Edwards, J.H. & Thompson, E.A. (Eds). Academic Press, London; 221-235 (1997). 5. Dicks, J. Graphical tools for comparative genome analysis. Yeast 17, 6-15 (2000). 6. Dalton, T.P., Edwards, J.H., Evans, E.P., Lyon, M.F., Parkinson, S.P., Peters, J. & Searle, A.G. Chromosome maps of man and mouse. Clin, Genet. 20, 407-15 (1981). 7. Edwards, J.H. Comparative genome mapping in mammals. Curr. Opin. Genet. Dev. 4, 861-7 (1994). 8. Morton, N.E. Gene maps and location databases. Ann. Hum. Genet. 55, 235-241 (1991). 9. Morton, N.E., Collins, A., Lawrence, S. & Shields, D.C. Algorithms for a location database. Ann. Hum. Genet. 56, 223-232 (1992). 10. Liao, W., Collins, A., Hobbs, M., Khatkar, MS., Luo, J. & Nicholas, F.W. A comparative location database (CompLDB): map integration within and between species. Mamm. Genome 18, 287-299 (2007).