The effect of prematurity on normal developmental programs of white and gray matter as evaluated with magnetic resonance imaging indicates global changes in white and gray matter with functional implications. We have previously identified an association between lipids and diffusion tensor imaging features in preterm infants, both through a candidate gene approach and a data-driven statistical genetics method. Here we apply a penalized linear regression model, the graph-guided group lasso (GGGL), that can utilize prior knowledge and select single nucleotide polymorphisms (SNPs) within functionally related genes associated with the trait. GGGL incorporates prior information from SNP-gene mapping as well as from the gene functional interaction network to guide variable selection.
|Title of host publication||Imaging Genetics|
|Number of pages||15|
|Publication status||Published - 26 Sept 2017|
- Brain development
- Machine learning
- Magnetic resonance imaging
- Sparse regression