Section 15 — Human Genetics 



strate clinical heterogeneity which suggests that 

 all cases may not be of the same genetic etiology. 

 Some have been separated with discriminate 

 function analysis of clinical data or by the de- 

 monstration of distinct biochemical differences 

 within diagnoses. These methods have failed 

 to define homogeneous groupings within other 

 diseases, including diabetes mellitus and cystic 

 fibrosis. On examination of the frequencies 

 of occurrence of a given disease in relatives of an 

 affected sibship an unbiased estimate of the 

 allelic frequency in the population may be ob- 

 tained on the assumption that homozygosity at 

 a single locus is responsible for all cases. The 

 population incidence estimated under this as- 

 sumption will be less than the observed popula- 

 tion incidence if the same gene locus is not segre- 

 gating in all observed families. 



Differences in the survival of affected indi- 

 viduals, age of onset and parental genotypes 

 require that different models be developed for 

 diabetes mellitus and cystic fibrosis on the as- 

 sumption that both of these are autosomal re- 

 cessive traits. The problem of reliable sampling 

 in cystic fibrosis families may lead to gross bias 

 in the fitting of the theoretical model. A method 

 is developed to obtain an approximation to 

 the population allelic frequency which deviates 

 by only 0.5 per cent from that value expected 

 under the theoretical model. 



From study of 754 cousins of cystic fibrosis 

 probands a population incidence of 1/3824 is 

 obtained. The similarity of this estimate to that 

 obtained from an epidemiologic study of a 

 comparable population (1/3700) suggests that 

 only one gene locus is involved in the etiology 

 of cystic fibrosis. 



15.5. (D.) The Application of Automatic Record 

 Linking Procedures. J. H. Edwards, I. M. 

 Leck, T. McKeown, and R.G. Record (Bir- 

 mingham, Great Britain). 



The application of automatic record linking 

 procedures and other methods dependent on 

 advanced computational methods to the Birm- 

 ingham data. These include detailed notifica- 

 tions of all births, with special reference to mal- 

 formations, and extensive linking of other 

 data. Births are identified by both mother and 

 child and sibships formed automatically as they 

 arise. 



cordant as to a given disease is always very im- 

 portant for Clinical Genetics, especially when 

 it is monozygotic. In Clinical Genetics, isolated 

 pairs occur to the physician, and the diagnosis 

 must therefore be based on intra-pair compari- 

 son. The statistical treatment of the case must 

 consider the fact that a concordant monozygotic 

 twin pair represents a statistical universe rather 

 than a statistical unit. This entails taking into 

 account many traits concerning the time of onset, 

 the symptoms, development, result of treat- 

 ment, etc. Suggestions are made for the statis- 

 tical treatment of such cases. 



15.7. First Findings on the Behaviour of Red Blood 

 Cells from MZ and DZ Twin Pairs in Isonia- 

 zide-fixation Tests (in Vitro). M. Bolognesi, 

 U. Fantoli and M. Milani-Comparetti 

 (Rome, Italy). 



An isoniazide-fixation test, described in the 

 paper, was carried out in vitro on red blood cells 

 from 30 female twin pairs, in order to investigate 

 its usefulness in zygosity determination. The 

 zygosity classification of the material was based 

 on several criteria of similarity including blood 

 groups (ABO, NM, Rh-Hr). 



The material was divided in two groups, res- 

 pectively under and above 16 years of age (to 

 guard against the higher frequency of fast inacti- 

 vators among individuals under 15, denounced 

 by Sunhara). 



Several tables contain the findings and statisti- 

 cal analyses; in particular Table V indicates 

 that only the variability due to zygosity exeeds 

 Fisher's F, being ascribable to chance only 

 with one probability out of 1000. The authors 

 conclude that the factor responsible for individual 

 behaviour in the test as applied seems to be under 

 genetic control, thus making it potentially useful 

 as a further test in zygosity determination. Fur- 

 ther research tends to: 



— determine the correlation between be- 

 haviour in this test and the fast or slow isoniazide 

 inactivator trait in vivo. 



— establish the limits of individual and in- 

 tra-individual variability for methodological codi- 

 fication. 



— extend the study to a wider sample, ap- 

 plying Kallman's twin family method to identify 

 the mode of transmission of the trait. 



15.6. On the Statistical Significance of One Pair of 

 Monozygotic Twins in Clinical Genetics. Luigi 

 Gedda (Rome, Italy). 



A case of one twin pair concordant or dis- 



15.8. Blood-groups and Haptoglobines by 200 

 Twin-pairs. Hans Niermann (Miinster, Ger- 

 many). 



For the diagnosis of twins it can be important 



272 



