Accelerated estimation and permutation inference for ACE modeling

Abstract

There are a wealth of tools for fitting linear models at each location in the brain in neuroimaging analysis, and a wealth of genetic tools for estimating heritability for a small number of phenotypes. But there remains a need for computationally efficient neuroimaging genetic tools that can conduct analyses at the brain‐wide scale. Here we present a simple method for heritability estimation on twins that replaces a variance component model‐which requires iterative optimisation‐with a (noniterative) linear regression model, by transforming data to squared twin‐pair differences. We demonstrate that the method has comparable bias, mean squared error, false positive risk, and power to best practice maximum‐likelihood‐based methods, while requiring a small fraction of the computation time. Combined with permutation, we call this approach “Accelerated Permutation Inference for the ACE Model (APACE)” where ACE refers to the additive genetic (A) effects, and common (C), and unique (E) environmental influences on the trait. We show how the use of spatial statistics like cluster size can dramatically improve power, and illustrate the method on a heritability analysis of an fMRI working memory dataset.

Short-channel regression in functional near-infrared spectroscopy is more effective when considering heterogeneous scalp hemodynamics

“…Non-negative least squares 48 – 50 was applied for the estimation of to reduce the risk of overfitting when including additional regressors. Non-negative least squares (MATLAB: lsqnonneg) iteratively minimizes the least-squared error between observation and the expected values by setting negative values to 0.…”

Genetic influences on brain activation and large-scale functional connectivity during nociceptive processing: a twin study

“…h 2 is equal to A/(A+C+E). In the present study, we computed heritability using the APACE software package (Accelerated Permutation Inference for the ACE model) (Chen et al, 2019). APACE uses a non-iterative linear regression-based method based on squared twin-pair differences, with permutation-based multiple testing correction to control the family-wise error rate.…”

“…For the mass-univariate analysis, for each first-level contrast described above, we used the Neurologic Pain Signature as a priori template for regions in which to test for significant differences in genetic influences between twin groups (Wager et al, 2013). The number of permutations was set to 1000 and we used the cluster-based inference in the APACE (Accelerated Permutation Inference for ACE models) software package (Chen et al, 2019) with cluster-forming threshold set to p < 0.05 based on the parametric likelihood ratio null-distribution. We additionally computed an estimate of the genetic influence of choice of threshold for the electrical stimulation using the mets package (Holst, Scheike, & Hjelmborg, 2016; Scheike, Holst, & Hjelmborg, 2014) implemented in R (R Core Team, 2017).…”

Towards understanding interindividual differences in cortical morphological brain networks

“…In this study, the heritability was estimated for morphological connectivity between each pair of regions for each type of single-layer morphological brain networks with the APACE package (Chen, Formisano et al 2019).…”

“…After identifying communities or modules embedded in the multiplex morphological brain networks, intra-and inter-module morphological connectivity were extracted and used to account for interindividual variance in multiple behavior and cognition domains with a multivariate variance component model (Ge, Reuter et al 2016, Liegeois, Li et al 2019, predict individual behavioral and cognitive performance with a multivariate brain basis set modeling method (Sripada, Angstadt et al 2019), and identify individuals using a network matching method (Finn, Shen et al 2015). Finally, an ACE model (Chen, Formisano et al 2019) was used to examine the extent to which morphological brain networks were under genetic control. We hypothesized that morphological brain networks 1) can account for interindividual variance in behavior and cognition; 2) can predict individual behavior and cognition;…”