Medical pedigrees: Typography and interfaces

We discuss the history of medical pedigrees and its importance for the design of the modern pedigree drawing software. 1 Writing on the wall The participants of TUG 2009 could see on the wall of Room 126, DeBartolo Hall, University of Notre Dame, a poster 23 feet wide. This poster depicted a pedigree of one of the authors of this paper (Leila). The pedigree included 13 generations (since the beginning of the 17 century), about three hundred persons in total. It was created in the following way: Leila did research in the National Archives of Bashkir Republic, interviewed her aunt and other relatives and deciphered the notes about family history dictated by her grandfather at mid-1980s. She put the data into an Excel file and periodically emailed the updated file to Boris, who then run it through a custom-made Makefile, getting as the result a PDF file. This file was e-mailed back. The final version was printed on two 16 foot sheets, which were taped together by Leila and her cousin in Ufa. This project became possible due to the suite of programs for automatic drawing of medical pedigrees we have been developing since 2005 (Veytsman and Akhmadeeva, 2007a; Veytsman and Akhmadeeva, 2007c). Their previous versions were discussed at TEX meetings (Veytsman and Akhmadeeva, 2006; Veytsman and Akhmadeeva, 2007b; Veytsman and Akhmadeeva, 2008) as well as conferences on genetics (Akhmadeeva, 2007). Now that the programs are no longer at the proof of concept level, we became more interested in improving them and making them of practical use for physicians, other health care specialists, geneticists, researchers, educators, and students. There are basically two ways to improve a typesetting program. First, we can improve its output, making the result more beautiful and clear. Second, we can improve its interface, making it more convenient, user-friendly and adding new features. In this paper we discuss both these approaches. 2 A (very) short history of pedigree typography A usual approach to beautiful typography is the study of history. The aesthetics of a printed page is based on the tradition. Thus it was natural for us to turn to the history of pedigrees in our quest to make them beautiful. We were fortunate to find the seminal paper (Resta, 1993) discussing just that topic. Resta unearthed many early pedigrees, and this section is based primarily on his research. One of the biggest surprises for us was the fact that pedigree drawing is a relatively modern art, started as late as the middle of the 19 century. One would think that people drew genealogical trees for ages, and thus pedigrees should have long history with venerable traditions. Still, this is not the case. While both genealogical trees and pedigrees are a graphical depiction of a family history, the goals of these two are completely different. A genealogical tree is intended to show the lineage of an individual, while a pedigree is used to show genetic traits. This difference in goals leads to a considerable difference in results. As an illustration of this difference, we show in Figure 1 a very nice genealogical tree created with PSTricks (Coustillas and Girou, 2004). The corresponding pedigree is shown in Figure 2. Note that the data of Figure 1 are actually not sufficient for the medical analysis, so we guessed that the marriage of Paul Joannon and Laure de Mortillet was childless. We also could not guess the relationship between the first and the second husbands of Élise Vidal; were they brothers? an uncle and a nephew?—so we decided against putting them on the chart. Even a casual glance at these two pictures shows the difference between the presentations. Clearly, the typographical traditions of genealogical trees is not very helpful for pedigrees. One of the first known medical pedigrees was published by Earle in the middle of the 19 century. He studied cases of color blindness in his own family. The pedigree included only females. Earle found out that his printer used to typeset musical sheets, so taking musical notes for symbols was a natural decision. Unaffected females were represented by half note symbols, while affected ones were represented by quarter-notes. Figure 3 shows the result. This nomenclature was apparently not used by anybody but Earle himself. Other scientists chose other ways to represent pedigrees. An influential researcher of the latter part of the century, Francis Galton, in his books and papers published in 1869– 1904 dispensed with drawing pedigrees, and used intricate tables instead (Resta, 1993). The rediscovery of Mendel’s seminal works on plant genetics might be the cause of the adoption of the symbols of Mars (Ä) and Venus (Ã) for males and females. A typical (hand-drawn) pedigree using this nomenclature is shown in Figure 4. It was Medical pedigrees: Typography and interfaces 228 TUGboat, Volume 30 (2009), No. 2 Affinité d’Élise Vidal avec Sabine Vigière d’Anval, sœur de Paul et avec Laure de Mortillet, sœur de la fiancée de Paul