FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W1517590395> ?p ?o ?g. }

• W1517590395 abstract "One of the major problems of MACE is the computation of the genetic correlations between countries. An advantage of structural models is the option to reduce the number of parameters to estimate. The structural model used here defines the genetic correlations between 2 countries as an exponential function of the Euclidian distance between them. In this structural model, (k+1) countries can be represented in a k-dimensional space. The reduction of the number of dimensions of the space allows to reduce the number of parameters. For example, in a 3-dimensional space only 72 coordinates need to be estimated to compute 325 genetic correlations between 26 countries. This structural model was successfully tested on simulated data, and on the current genetic correlations matrix used by INTERBULL. The first aim of the present study was to compare results of the structural model used on international data with results of an unstructured model. The second .aim was to study the possible use of the coordinates of different structural models related to the same space, to calculate directly genetic correlations, without having to estimate them. The third aim was to analyze the influence of the choice of the axes countries. Deregressed national breeding values used for Holstein milk yield international evaluation of February 2003 were analyzed. Several subsets of countries were considered. The structural model and a classical model were applied to estimate the genetic correlations between countries, using an AI-REML algorithm implemented in the AIREMLF90 program. Countries chosen to define axes in the structural model were based on the results from a previous study that applied a cluster analysis: The Netherlands as centre of the space, USA to define the I axis, New Zealand for the 2 axis and Hungary for the 3 axis. Genetic correlations estimated with the structural model were very close to those estimated with the classical model. Larger differences (e.g., 0.06 for the correlation between Denmark and New Zealand) concerned the least connected of the countries considered. The standard errors of genetic correlations ranged from 0.006 to 0.058 depending on the amount of genetic ties. The -210gL was slightly higher for the structural model than for the classical model, but Bayesian information criterion favoured the structural model because of the lower number of parameters. Combining the coordinates obtained from different structural models related to the same space gave similar genetic correlations (differences were lower than the standard errors) and similar -210gL compared to the joint analyses of the countries. By this way, the number of runs needed to estimate all the genetic correlations can be reduced drastically. The change of some of the axes countries shows that the estimated genetic correlations were more precise when the volume defined by the axes countries was large. Determination of the optimal number of countries to define the axes and the choice of the axes countries needs to be investigated further, using e.g. information criteria and volume of the space defined by axes countries." @default.
• W1517590395 created "2016-06-24" @default.
• W1517590395 creator A5043481359 @default.
• W1517590395 creator A5072823605 @default.
• W1517590395 creator A5073249281 @default.
• W1517590395 date "2003-01-01" @default.
• W1517590395 modified "2023-10-18" @default.
• W1517590395 title "Application of a structural model to estimate genetic correlations between countries." @default.
• W1517590395 cites W1579233395 @default.
• W1517590395 cites W1604642794 @default.
• W1517590395 cites W1707134449 @default.
• W1517590395 cites W1973930407 @default.
• W1517590395 cites W1974594847 @default.
• W1517590395 cites W1993849755 @default.
• W1517590395 cites W1994552776 @default.
• W1517590395 cites W2010907879 @default.
• W1517590395 cites W2032704718 @default.
• W1517590395 cites W2040161470 @default.
• W1517590395 cites W2053115400 @default.
• W1517590395 cites W2059217163 @default.
• W1517590395 cites W2065319808 @default.
• W1517590395 cites W2068484625 @default.
• W1517590395 cites W2070284821 @default.
• W1517590395 cites W2079740932 @default.
• W1517590395 cites W2133440243 @default.
• W1517590395 cites W2157019764 @default.
• W1517590395 cites W2174812595 @default.
• W1517590395 cites W2270334462 @default.
• W1517590395 cites W2787936211 @default.
• W1517590395 hasPublicationYear "2003" @default.
• W1517590395 type Work @default.
• W1517590395 sameAs 1517590395 @default.
• W1517590395 citedByCount "4" @default.
• W1517590395 crossrefType "proceedings-article" @default.
• W1517590395 hasAuthorship W1517590395A5043481359 @default.
• W1517590395 hasAuthorship W1517590395A5072823605 @default.
• W1517590395 hasAuthorship W1517590395A5073249281 @default.
• W1517590395 hasConcept C149782125 @default.
• W1517590395 hasConcept C162324750 @default.
• W1517590395 hasConcept C41008148 @default.
• W1517590395 hasConceptScore W1517590395C149782125 @default.
• W1517590395 hasConceptScore W1517590395C162324750 @default.
• W1517590395 hasConceptScore W1517590395C41008148 @default.
• W1517590395 hasLocation W15175903951 @default.
• W1517590395 hasOpenAccess W1517590395 @default.
• W1517590395 hasPrimaryLocation W15175903951 @default.
• W1517590395 hasRelatedWork W2130043461 @default.
• W1517590395 hasRelatedWork W2350741829 @default.
• W1517590395 hasRelatedWork W2358668433 @default.
• W1517590395 hasRelatedWork W2376932109 @default.
• W1517590395 hasRelatedWork W2382290278 @default.
• W1517590395 hasRelatedWork W2390279801 @default.
• W1517590395 hasRelatedWork W2748952813 @default.
• W1517590395 hasRelatedWork W2899084033 @default.
• W1517590395 hasRelatedWork W3004735627 @default.
• W1517590395 hasRelatedWork W2530322880 @default.
• W1517590395 isParatext "false" @default.
• W1517590395 isRetracted "false" @default.
• W1517590395 magId "1517590395" @default.
• W1517590395 workType "article" @default.