Publications by Michel G. Nivard & Perline Demange
Overcomplete factor analysis in MCMSEM
Overcomplete factor analysis in MCMSEM Author Michel Nivard Published September 25, 2024 This demo explores the integration of overcomplete Independent Component Analysis (ICA) and Structural Equation Modeling (SEM) using the MCMSEM package in R. Traditional methods like PCA, ICA, and SEM typically have to assume there are fewer (or equal) la...
13139 sym Python (17194 sym/35 pcs) 5 img
Estimating Returns to Education with MCMSEM
Estimating Returns to Education with MCMSEM Author Michel G Nivard MCMSEM Multi Co-Moment Structural equation models (MCMSEMs) are structural equation models that consider covariance, coskewness and co-kurtosis between variables. This means there is sufficient information to estimate models based on observational data, where we do not know the ...
8618 sym Python (6765 sym/17 pcs) 4 img
1%
one_procent Author Michel Nivard A rare (0.5%) genetic variant with R2 1%, but i super predictive for carriers. Please enable JavaScript to experience the dynamic code cell content on this page. But it’ll look super predictive for carriers! Please enable JavaScript to experience the dynamic code cell content on this page....
341 sym
Are cousins informative
Are cousins informative? Author Michel Nivard Can cousins inform us about the alleles we didnt inherint form our parents? This is simualiton code to show that cousins inform us about the alleles we did not inherint form our parents. set basic parameters, 1000 people 1 allele with an 25% minor allele frequency n <- 1000 maf <- .25 draw the g...
680 sym
GWAS-by_subtraction tutorial
Performing a GWAS-by-subtraction in GenomicSEM. The preprint by Demange et al. (here on BioRxiv) uses GenomicSEM and the input GWAS of educational attainment (EA) and cognitive performance (CP) to derive two new GWASes: cognitive (Cog) and non-cognitive (NonCog) contributions to educational attainment. We feel this is a very valuable analysis as ...
7296 sym R (4246 sym/20 pcs) 1 img
HDL in GenomicSEM
HDL in GenomicSEM High-definition likelihood (HDL) inference is an exciting new method to estimate heritability, and genetic correlation with higher precision then LD score regression (LDSC). LDSC models the relation between the covariance of a SNP’s Z statistics for pairs of traits and that SNP’s LD score. The HDL extends the LDSC method by ...
4820 sym R (4807 sym/14 pcs) 1 tbl
Twin models in Lavaan
Introduction Resemblance between relatives There is a long history of questioning the nature of the resemblance between family members, specifically sibling and even more specifically twins. Reasons that are commonly put forward for these similarities are the obvious genetic similarity, similarity in socio-economic position, the childhood enviro...
22691 sym R (54032 sym/42 pcs) 5 img 1 tbl