Structural equation mixture modeling and their application in analysis of SNP
Yanbo Zhang
Abstract
Yanbo Zhang
Abstract
To analyze SNP data of GAW17 by Structural equation mixture modeling (SEMM), and to provide a new method for the study of genetic statistic . The data is provided by GAW17, it contains 697 individual, 22 autonomic tens of thousands of SNP and the SNP simulated 697 individual trait characteristics. In this study, randomly selected the four SNP from chromosome 1 and three quantitative traits as a research variable,which were analysised by latent class and mixed structural equation modeling. According to four SNP data, the crowd was divided into three potential categories, each category probability were 0.53, 0.34, 0.13. Factors mean Q of latent class 1, 2 and 3 are -4.029, -2.052 and 0.We knew that factor mean of latent class 1, 2 are lower than 3 (0.001). So we have reasons to think that structural equation mixed modeling integrated the structural equation modeling and latent class modeling thoughts, formed its own advantage,which can be used for processing classification latent variable and continuous latent variable data.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
To analyze SNP data of GAW17 by Structural equation mixture modeling (SEMM), and to provide a new method for the study of genetic statistic . The data is provided by GAW17, it contains 697 individual, 22 autonomic tens of thousands of SNP and the SNP simulated 697 individual trait characteristics. In this study, randomly selected the four SNP from chromosome 1 and three quantitative traits as a research variable,which were analysised by latent class and mixed structural equation modeling. According to four SNP data, the crowd was divided into three potential categories, each category probability were 0.53, 0.34, 0.13. Factors mean Q of latent class 1, 2 and 3 are -4.029, -2.052 and 0.We knew that factor mean of latent class 1, 2 are lower than 3 (0.001). So we have reasons to think that structural equation mixed modeling integrated the structural equation modeling and latent class modeling thoughts, formed its own advantage,which can be used for processing classification latent variable and continuous latent variable data.
Key concepts: Structural equation modeling, Latent class model, Latent variable, Latent variable model, SNP, Statistic, Statistics, Class (philosophy)