In a previous study we showed that prediction tools are useful to select single-nucleotide polymorphisms (SNPs) which potentially affect phenotype and therefore guide genotyping in association studies, thus saving time and money. Here we use the recently available RNA cross-link immunoprecipitation data to analyze several genes involved in psychiatric disorders and show which disease-associated SNPs can affect the splicing process by altering splicing factor binding sites. We point out the importance of using cross-link immunoprecipitation data in psychiatry to refine the SNP selection methods, to explain the association found and to plan molecular investigations.