Genomic experiments encapsulate such rich information that specific analysis tools are needed to take full advantage of the available data. Recently, more attention has been devoted to developing tools to identify the association between genomic variations and phenotypic traits. This paper focuses on understanding an approach called parallel independent component analysis (ICA), while applied to massive microarray data. Parallel ICA is a multivariate approach which can process genotypic and phenotypic data in parallel while emphasizing the relationship between data types. We simulated single nucleotide polymorphism (SNP) array data and functional magnetic resonance imaging data, generating hidden connections between the two modalities. The performance of parallel ICA was compared with a correlation test for a single hypothesis. Our evaluation shows that parallel ICA, in general, is able to extract more accurately the components and connections than the correlation test, in particular for weak linkages. Results also indicate that the ratio of sample size to SNP size should be at least 0.02. However, when the data have a low odds ratio or cases vs. controls ratio, the correlation test provides results reliably, though with lower accuracy.