Biomedical subjects
Publications and source records attributed to S Yee.
Numerical analysis is applied to physical assignments of loci, providing point estimates, an LRO confidence interval, and a chi(2) test of consistency whether there is a smallest region of overlap (SRO) or not. Results are given for two examples and summarized for 81 loci in man.
A general linear model is presented here for biological and cultural inheritance involving ten parameters to be estimated from 16 correlations in nuclear families, providing ample degrees of freedom to test goodness of fit. Applied to six lipoprotein traits the model fits acceptably to all, although there is evidence of transient maternal effects for cholesterol and lipemia. Genetic heritability in children ranges from 0.175 for triglyceride to 0.562 for total cholesterol. Cultural heritability in children ranges from 0.012 for VLDL to 0.149 for HDL-cholesterol.
Thirteen loci are mapped on chromosome 1 from genetic evidence. The maximum likelihood map presented permits confirmation that Scianna (SC) and a fourteenth locus, phenylketonuria (PKU), are on chromosome 1, although the location of the latter on the PGM1-AMY segment is uncertain. Eight other controversial genetic assignments are rejected, providing a practical demonstration of the resolution which maximum likelihood theory brings to mapping.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Maximum likelihood methods are presented to test for the relations between causes and effects in linear path diagrams, without assuming that estimates of causes are free of error. Causal analysis is illustrated by published data of the Equal Educational Opportunity Survey, which show that American schools do not significantly modify socioeconomic differences in academic performance and that little of the observed racial difference in academic performance is causal. For two races differing by 15 IQ points, the differential if social class were randomized would be only about 3 points. The principle is stressed that a racial effect in a causal system may be environmental and that its etiology can be studied only by analysis of family resemblance in hybrid populations.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Kinship of Barra (Outer Hebrides) is bioassayed as 0.0096 relative to northwestern Europe, in reasonable agreement with prediction of 0.008 relative to Britain. The exponential decline of kinship with distance is similar to Scandinavia. Kinship of larger populations is consistent with predictions from isolates. Kinship is largely due to local drift rather than Norse admixture, the estimate of which is obscured by drift and appears highly unreliable. Other populations in northwest Europe, including Iceland, Lewis and Orkney, give estimates of kinship which reflect drift and geography, but also do not provide reliable estimates of admixture. Bioassay of kinship from gene frequencies gives observations to be explained in terms of migration and drift, but contributes nothing to the question of whether polymorphism in the species is maintained by neutral mutation or selection. Although the latter is on various grounds more likely, no calculations based on gene frequency distribution can provide critical evidence.
By fitting compounds beta distributions to chiasma frequencies the physical map obtained from banded chromosomes has been converted into a chiasma map giving the distribution of observed chiasmata in relation to several hundred cytological bands, assuming proportionality of mitotic and meiotic chromosomes. This is a genetic map if there is a precise correspondence between sites of chiasmata and crossing-over. However, if there is appreciable preanaphase movement of chasmata, then the chiasma map is a serious distortion of the genetic map. Predictions from the chiasma map can be confirmed or refuted only by genetic evidence for which the estimates of this paper serve as initial values to begin maximum likelihood iteration.
Assuming a perfect correspondence between the site of crossing-over and an observed chiasma, data on meiosis in the human male are used to estimate a mapping parameter which on average turns out to be intermediate between the Kosambi and Carter-Falconer values, but smaller for acrocentrics. A table is given for converting recombination frequencies to map distances.
In a sample of nearly 8,000 Japanese males, the distributions of casual cholesterol and triglyceride, hematocrit, glucose, uric acid, diastolic blood pressure, and weight (covariance adjusted) could not be normalized by a power transform and were significantly better fitted by a mixture of distributions. The evidence for admixture was nonsignificant for systolic blood pressure, significant but unimpressive for height and weight, and strong for the remaining variables. The minor component corresponded to high values, in low frequency except for triglyceride and glucose. These results favor an interpretation of elevated levels in terms of distinct entities, genetic or environmental or both, rather than cumulative small effects only. These entities appear to be megaphenic (i.e., with effects exceeding one phenotypic standard deviation). Consequences of this hypothesis are discussed.
The chiasma map of D. melanogaster is inferred from the genetic map, giving correspondences between physical and genetic locations. Crossing-over is reduced near the centromere and telomeres. If chiasmata occur preferentially near the telomere they must be distal to genetic loci. A precept of Bridges and Morgan is discussed, which Drosophila genetics neglected but chromosome mapping in other organisms should follow.
Historical demography, surname concordance (isonymy), migration, and genealogy give a consistent description of population structure. The census size has averaged about 1400 over the last five centuries. Conjoined with an effective migration rate of 3-05 per generation as estimated by three different methods, this gives an evolutionary size of 638, random kinship of 0-008 and inbreeding of 0-007 relative to the rest of Britain. The population structure of Barra is similar to other British isolates in the recent past, but an order of magnitude less inbred than slash-and-burn agriculturalists and Pacific Islanders. Some consequences for rare genes and polymorphisms are discussed.
Analysis of family resemblance is developed in terms of three genetic parameters, six parameters for cultural inheritance, and one parameter for an index estimating family environment. With efficient use of nuclear families the model is fully determinate. Other biological and social relationship provide additional degrees of freedom for testing goodness of fit. Performance of the model is satisfactory on simulated data with extreme gene-environment interaction. Applied to a large body of published data on I.Q., neither genetic assortative mating nor gene-environment covariance is significant by a likelihood ratio test, but heritability is less and cultural inheritance is greater for adults than children. Whereas family resemblance of children is largely genetic, for adults it is largely due to their childhood environments, presumably acting on occupational aspirations. Further resolution is more likely to come from nuclear families than from the rare relationships that were favored by classical human genetics.
A likelihood ratio test is given for distinguishing skewness from commingled distributions, using a power transform to remove skewness appropriately for each of the alternatives tested. The alternative hypotheses postulate that the transformed data are from one normal or a mixture of two or three normal homoscedastic distributions. Since each mixture has unique asymmetry, skewness is estimated simultaneously with the means, proportions and variance of components. Commingling cannot be rigorously proven in this way, as some other transform may provide a better approximation to normality. However, the error of asserting admixture whenever there is skewness has been avoided, and estimates of admixture parameters provide a basis for more conclusive tests in relatives or other populations. Two examples are given, one in which adjustment for skeweness left evidence of commingling.
Sib correlations for height and weight decrease with absolute age difference. Parent-child correlations increase with age of child, with greater resemblance to the mother than the father. Estimates of the relative variance due to common environment are greater, and heritability estimates less, than for earlier studies. Heritability is less for adults than children. There is no significant major locus for height or weight.