Human gene for torsion dystonia located on chromosome 9q32-q34.

Torsion dystonia is a movement disorder of unknown etiology characterized by loss of control of voluntary movements appearing as sustained muscle contractions and/or abnormal postures. Dystonic movements can be caused by lesions in the basal ganglia, drugs, or gene defects. Several hereditary forms have been described, most of which have autosomal dominant transmission with variable expressivity. In the Ashkenazi Jewish population the defective gene frequency is about 1/10,000. Here, linkage analysis using polymorphic DNA and protein markers has been used to locate a gene responsible for susceptibility to dystonia in a large, non-Jewish kinship. Affected members of this family have a clinical syndrome similar to that found in the Jewish population. This dystonia gene (ITD1) shows tight linkage with the gene encoding gelsolin, an actin binding protein, and appears by multipoint linkage analysis to lie in the q32-q34 region of chromosome 9 between ABO and D9S26, a region that also contains the locus for dopamine-beta-hydroxylase.

Correlations in multi-locus genetic systems: a mathematical model.

A spin model with parallels to a multi-locus genetic model is presented which makes it possible to calculate the correlation between allelic states at any two loci in a population at equilibrium. The main features are: (1) The decay of correlation with distance may be expressed essentially as a linear combination of two exponentials, one of which dominates when the two loci are sufficiently far apart. (2) The correlation between two loci a specified distance apart is increased as the number of loci in the entire system increases. The results are compared with those of other theoretical models and discussed in the light of available experimental data. Possible ways of generalizing the model are outlined. However, additional experimental data is clearly needed to indicate the genetic relevance of work of this nature.

Characteristics of a multiplex IDDM sample: unexplained differences with other samples.

Characteristics of a multiplex sample of families with insulin-dependent diabetes mellitus (IDDM) are studied and contrasted with similar characteristics in other, more conventionally sampled data sets. Some characteristics remain consistent with earlier observations including the high frequency of human leukocyte antigen (HLA) DR3,4 in affected individuals and the greater than expected percentage of HLA haplotype sharing among affected sib pairs. In other respects, however, differences are seen between this sample and others. "Control" haplotypes, i.e., those not transmitted to the first affected offspring, had a higher frequency of DR3 and DR4 than expected, and a rather high frequency of affected parents was observed. Differences between the first affected and later affected offspring reported in other samples were absent from these families. No effect of the sampling scheme and the resulting distribution of parental phenotypes could be shown to explain this difference.

A two-susceptibility-allele model for genetic diseases and associated marker loci: differences and similarities to a one-s-allele model.

A model for genetic diseases and associated markers is defined where two distinct susceptibility alleles are possible, each associated with a different marker allele. Marker genotype distributions in a disease population are then expressed in terms of haplotype frequencies and penetrance parameters. It is shown that, if the heterozygote with two different disease alleles has a higher penetrance than the two disease homogzygotes, the observed to 'Hardy-Weinberg-expected' ratio of associated marker genotypes (the alpha/beta ratio of Falk, Mendell & Rubinstein, 1983) will always be greater than or equal to one. When all disease penetrances are equal, the model becomes indistinguishable from a recessive one-s-allele model with alpha/beta = 1. Application of these observations to several data sets for insulin dependent diabetes mellitus suggests the possibility that different marker genotype distributions in different samples may be due to different penetrances of the disease genotypes in the samples. If a particular environment causes the heterozygote disease genotype (with two different disease alleles) to have the highest penetrance, the marker genotype distribution would be compatible with the 2-s-allele model. In other environments where the three disease genotypes have essentially equal penetrances, the marker distribution would be compatible with the 1-s-allele model.

Preliminary ordering of multiple linked loci using pairwise linkage data.

A method is presented for the preliminary ordering of loci on a chromosome using pairwise linkage data. The method is based on the biologically reasonable assumption that the "true" order of a set of linked loci will be the one that minimizes the total length of the chromosome segment. Here the "length" is defined as the sum of adjacent recombination fractions. The method searches for the optimal order, represented by a minimum distance map (MDMAP), even when it is not possible to examine the n!/2 possible distinct orders for n loci. A computerized approach, using the simulated annealing algorithm of Kirkpatrick et al. [1983], forms the basis of the method. It can be applied to data from radiation hybrid experiments as well as that from conventional family linkage studies. The technique is applied to several sets of published data to illustrate how it performs in practice. The advantages and the disadvantages of the method are discussed so that it will be clear under what conditions it is likely to work well. When data sets are "complete," in the sense that all possible pairwise recombination fractions have estimates, and when no large clusters of extremely tightly linked loci are present, the method produces ordered sets of loci that agree well with those generated by other, more complex methods. Any discrepancies that occur are likely to be with respect to the orientation of nearest-neighbor loci, where relative order cannot be reliably established by any method. The method thus provides a simple, rapid means of obtaining a preliminary order for a set of loci known to be in the same linkage group.

A simple method for ordering loci using data from radiation hybrids.

A method is presented for ordering loci on a chromosome based on data generated from radiation hybrids. All loci are tabulated as being present, absent, or not scored in a series of clones. Correlation coefficients are calculated for all pairs of loci indicating how often they are retained or lost together in the clones. On the assumption that a high positive correlation implies closely linked loci, a distance score, d, equal to one minus the correlation coefficient, is obtained for each locus pair and an order is generated that minimizes the sum of the adjacent distances [the MDMAP method of Falk ("Multipoint Mapping and Linkage Analysis Based upon Affected Pedigree Members: Genetic Analysis Workshop 6," pp. 17-22, A. R. Liss, New York, 1989)]. Two sets of data, with information on 13 and 16 loci mapped to chromosome 21q, have been ordered using this method. The results are in very good agreement with other ordering methods used on the same data and with physical mapping data.

Effect of genetic heterogeneity and assortative mating on linkage analysis: a simulation study.

Linkage studies of complex genetic traits raise questions about the effects of genetic heterogeneity and assortative mating on linkage analysis. To further understand these problems, I have simulated and analyzed family data for a complex genetic disease in which disease phenotype is determined by two unlinked disease loci. Two models were studied, a two-locus threshold model and a two-locus heterogeneity model. Information was generated for a marker locus linked to one of the disease-defining loci. Random-mating and assortative-mating samples were generated. Linkage analysis was then carried out by use of standard methods, under the assumptions of a single-locus disease trait and a random-mating population. Results were compared with those from analysis of a single-locus homogeneous trait in samples with the same levels of assortative mating as those considered for the two-locus traits. The results show that (1) introduction of assortative mating does not, in itself, markedly affect the estimate of the recombination fraction; (2) the power of the analysis, reflected in the LOD scores, is somewhat lower with assortative rather than random mating. Loss of power is greater with increasing levels of assortative mating; and (3) for a heterogeneous genetic disease, regardless of mating type, heterogeneity analysis permits more accurate estimate of the recombination fraction but may be of limited use in distinguishing which families belong to each homogeneous subset. These simulations also confirmed earlier observations that linkage to a disease "locus" can be detected even if the disease is incorrectly defined as a single-locus (homogeneous) trait, although the estimated recombination fraction will be significantly greater than the true recombination fraction between the linked disease-defining locus and the marker locus.

Locus ordering based on crossover information in family haplotypes: application of a "minimum break" algorithm.

A "minimum break" method, previously applied to radiation hybrid data, is applied to the Genetic Analysis Workshop 12 simulated family data to assess the method's feasibility for ordering dense markers that have been typed in a set of pedigrees of three or more generations. Ordering is based on minimizing the total number of crossovers ("breaks") in a set of haplotypes where allele origin can be assigned to each specific haplotype, but locus order is unknown. It is shown that, for average locus spacing of about 2 cM and a set of 300 to 600 haplotypes, locus ordering can be reliably determined for sets of at least 50 loci. However, in the absence of crossovers between some pairs (sets) of loci, no unique order can be identified and sets of orders that are equally likely will be recovered. The method presented can be used to obtain reasonable sets of ordered loci that can then be used by other methods to refine the statistical confidence of the locus order.

Systematic search for disease loci for complex genetic traits: a study based on simulated population data.

Simulated family data were analyzed using one- and two-locus disease models to detect linkage. Regions of interest, found on chromosomes 3 and 5, were then further analyzed to look for evidence of locus interaction and/or genetic heterogeneity. Methods described by Falk [1993] were used to separate families into subsets likely to be genetically homogeneous. Based on the results, it was concluded that there were at least two distinct disease loci, one on chromosome 3 and one on chromosome 5, and that these loci were probably not interacting but were expressing two distinct forms of the disease. The identification of these loci was in agreement with the generating model. However, the analysis did not show any indication of a two-locus form of the disease or detect a disease locus on chromosome 1. This could be due to lack of power and/or too small a sample size for the method of analysis.

Haplotype relative risks: an easy reliable way to construct a proper control sample for risk calculations.

An alternative to Woolf's (1955) relative risk (RR) statistic is proposed for use in calculating the risk of disease in the presence of particular antigens or phenotypes. This alternative uses, as the control sample, the parental antigens or haplotypes not present in the affected child. The formulation of a haplotype relative risk (HRR) thus eliminates the problems of sampling from the same homogeneous population to form both the disease sample and an appropriate control. We show that, in families selected through a single affected individual, where transmission of the four parental haplotypes can be followed unambiguously, the mathematical expectation of the HRR is identical to that of the RR. Since the sample formed from the 'non-affected' parental haplotypes is clearly from the same population as the disease sample, the HRR thus provides a reliable alternative to the RR. A further advantage obtains when family data are being collected as part of a study since the control sample is then automatically contained in the family material. Data from studies of patients with insulin dependent diabetes mellitus (IDDM) are used to obtain an estimate of the risk to those with HLA antigens or phenotypes associated with IDDM using the HRR statistic. A comparison of the HRR's and RR's for these data is also presented.

The effect of association of genetic factors on relative risk.

We have examined the numerical properties of relative risk (RR) estimates under the conditions of gene frequency and genetic association (delta) that generally obtain for the different HLA loci. It is found that significant RR may be created solely by the existence of positive delta between the disease susceptibility gene and a specific allele at the marker (e.g. HLA) locus. It is shown further, that the value of RR is strongly affected by the gene frequencies and not only by the corresponding delta. If positive delta exists between one disease susceptibility gene and two HLA haplotypes, the RR estimates will usually be highest for the corresponding heterozygote, intermediate for the two homozygotes and lowest for the other carrier phenotypes. This model was found to be perfectly compatible with the situation encountered in the case of juvenile diabetes mellitus and shows that 'overdominance' is not required to account for the excessive RR found in some populations for certain HLA heterozygous classes.

Genetic loci ordering instability: an example.

In attempting to establish the order of genetic loci by constructing a map from pairwise linkage data, one assumes that the loci satisfy a linear-order relation. If the data utilized in the construction are not consistent with the linear-order assumption, then a very small change in the data may lead to a large qualitative change in the map. An example of such an instability is presented in this paper.

Effect of population associations and reduced penetrance on observed and expected genotype frequencies in a simple genetic model: application to HLA and insulin dependent diabetes mellitus.

The observed and Hardy--Weinberg-expected frequencies for (HLA) marker heterozygotes in a disease population are calculated where it is assumed that the disease is caused by either homozygosity or heterozygosity (with reduced pentrance) for a disease susceptibility allele at a disease locus, that allele being positively associated with both of the relevant alleles at the marker locus (the single susceptibility allele model of Svejgaard & Ryder, 1981). It is shown that the observed frequency is always less than or equal to the H-W expectation, with the (observed/expected) ratio decreasing as the degree of dominance increases.

Design of artificial neural network and its applications to the analysis of alcoholism data.

Artificial neural networks were applied to the alcoholism data to reveal nonlinear relationships between intermediate phenotypes, marker identity-by-descent sharing, and the affection status. A variable number of hidden units were considered to achieve a balance between the minimal mean-squared error and over-fitting of the data. The predictability of the affection status based on intermediate phenotype information (event-related potential 300, monoamine oxidase, and gender) was 65% to 75%, and sensitivity/specificity ranged around 50% to 80%. The IBD approach succeeded in identifying the same marker as previous studies, but also found additional peaks.

Using neural networks as an aid in the determination of disease status: comparison of clinical diagnosis to neural-network predictions in a pedigree with autosomal dominant limb-girdle muscular dystrophy.

Studies of the genetics of certain inherited diseases require expertise in the determination of disease status even for single-locus traits. For example, in the diagnosis of autosomal dominant limb-girdle muscular dystrophy (LGMD1A), it is not always possible to make a clear-cut determination of disease, because of variability in the diagnostic criteria, age at onset, and differential presentation of disease. Mapping such diseases is greatly simplified if the data present a homogeneous genetic trait and if disease status can be reliably determined. Here, we present an approach to determination of disease status, using methods of artificial neural-network analysis. The method entails "training" an artificial neural network, with input facts (based on diagnostic criteria) and related results (based on disease diagnosis). The network contains weight factors connecting input "neurons" to output "neurons," and these connections are adjusted until the network can reliably produce the appropriate outputs for the given input facts. The trained network can be "tested" with a second set of facts, in which the outcomes are known but not provided to the network, to see how well the training has worked. The method was applied to members of a pedigree with LGMD1A, now mapped to chromosome 5q. We used diagnostic criteria and disease status to train a neural network to classify individuals as "affected" or "not affected." The trained network reproduced the disease diagnosis of all individuals of known phenotype, with 98% reliability. This approach defined an appropriate choice of clinical factors for determination of disease status. Additionally, it provided insight into disease classification of those considered to have an "unknown" phenotype on the basis of standard clinical diagnostic methods.

Genetic Analysis Workshop II: summary.

The second Genetic Analysis Workshop was held October 30, 1983, at the American society of Human Genetics meetings in Norfolk, Virginia. Ten groups of investigators analyzed three sets of pedigrees generated by computer simulation. A summary of the workshop is presented. In this overall description by the workshop organizers, the rationale for the choice of problems and the characteristics of the three data sets are discussed, and the ten analyses are briefly summarized and compared. Following this general summary are brief descriptions of the analyses of each group of workshop participants, in alphabetical order by senior author.