Multivariate normal genetic models with a finite number of loci
Abstract
A genetic model due to Russell Lande is described. The model assumes a finite number of loci at each of which there are an infinite number of alleles whose effects on the phenotype are normally distributed. Analytic and numerical results using this model depend on the allele effects remaining multivariate normally distributed. This is almost never exactly true, but may often be a good approximation. Numerical results for several kinds of natural selection are discussed. A model involving overdominance is presented which seems to exhibit the FranklinLewontin crystallization effect. A model is presented of the maintenance of genetic variation by a cline along which there is linear change in the optimum phenotype under optimizing selection. The equilibrium has been found analytically for the case of an infinite cline. Remarkably, there is no linkage disequilibrium maintained at equilibrium in this case.
 Authors:
 Publication Date:
 Research Org.:
 Washington Univ., Seattle (USA). Dept. of Genetics
 OSTI Identifier:
 6315326
 Alternate Identifier(s):
 OSTI ID: 6315326
 Report Number(s):
 RLO2225/539
 DOE Contract Number:
 EY76S062225005
 Resource Type:
 Technical Report
 Country of Publication:
 United States
 Language:
 English
 Subject:
 59 BASIC BIOLOGICAL SCIENCES; GENETIC VARIABILITY; GENETICS; MATHEMATICAL MODELS; CHROMOSOMES; PHENOTYPE; BIOLOGICAL VARIABILITY; BIOLOGY 550400*  Genetics
Citation Formats
Felsenstein, J. Multivariate normal genetic models with a finite number of loci. United States: N. p., 1977.
Web.
Felsenstein, J. Multivariate normal genetic models with a finite number of loci. United States.
Felsenstein, J. 1977.
"Multivariate normal genetic models with a finite number of loci". United States.
doi:.
@article{osti_6315326,
title = {Multivariate normal genetic models with a finite number of loci},
author = {Felsenstein, J.},
abstractNote = {A genetic model due to Russell Lande is described. The model assumes a finite number of loci at each of which there are an infinite number of alleles whose effects on the phenotype are normally distributed. Analytic and numerical results using this model depend on the allele effects remaining multivariate normally distributed. This is almost never exactly true, but may often be a good approximation. Numerical results for several kinds of natural selection are discussed. A model involving overdominance is presented which seems to exhibit the FranklinLewontin crystallization effect. A model is presented of the maintenance of genetic variation by a cline along which there is linear change in the optimum phenotype under optimizing selection. The equilibrium has been found analytically for the case of an infinite cline. Remarkably, there is no linkage disequilibrium maintained at equilibrium in this case.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = 1977,
month = 1
}

Results are reported from studies on somatic and genetic effects of x radiation on tomato seed and pollen. The presence of some mechanism that affords protection or acts as a buffer against genetic injury was suggested when seeds were irradiated under partially hydrated conditions. The characteristics of radioinduced mutations are discussed. Data are included on the frequency and biochemical characteristics of mutations resulting from the irradiation of hydrated pollen. (C.H.)

Models for aromatics in gasolines based on near infrared spectroscopy, NIR, and multivariate calibration. Comparison with models for aromatics in kerosenes
The objective of this work was to find a model which could predict the percentage of aromatics in gasolines from a near infrared spectra (NIR) of the gasolines. To do this we use a new method developed by Vigerust et al. (1992) which is based on Multivariate Statistical Calibration of NIR spectra. A calibration model is developed based on a calibration set consisting of gasoline samples where both the NIR spectra and the percentage of aromatics are known. By using calibration models like this Vigerust et al. (1992) have shown that the time required for determining aromatics in kerosenes duringmore » 
Source apportionment of air pollution in China: Extending the usefulness of receptor modeling by combining multivariate and chemical mass balance models
The research explores the possibility of using a two step method of identifying and quantifying air pollution emissions in an urban environment. The procedures uses a mathematical model called Target Transformation Factor Analysis (TTFA) to estimate source profiles using ambient trace element air concentration data. A source profile is analogous to a fingerprint since it is unique to each source of air pollution. The profiles estimated by TTFA are then employed in a Chemical Mass Balance (CMB) source apportionment analysis for the air shed. Other known sources are estimated using source signatures from the literature. Applying the TTFA and CMBmore »