%0 Journal Article
%T Finding disease candidate genes by liquid association
%A Ker-Chau Li
%A Aarno Palotie
%A Shinsheng Yuan
%A Denis Bronnikov
%A Daniel Chen
%A Xuelian Wei
%A Oi-Wa Choi
%A Janna Saarela
%A Leena Peltonen
%J Genome Biology
%D 2007
%I BioMed Central
%R 10.1186/gb-2007-8-10-r205
%X Studies aiming to identify susceptibility genes in complex diseases have proceeded along two lines. The traditional candidate gene approach is limited by our ability to come up with a comprehensive list of biologically related genes. On the other hand, the 'hypothesis free' approach relies on genome-wide scans for disease loci, typically via linkage in exceptionally large families or via association in case control studies. Multiple sclerosis (MS), which is one of the most common neurologic disorders affecting young adults, is characterized by demyelination and reactive gliosis [1]. Analogous to many complex traits, genome scans in MS have identified numerous chromosomal loci often with only a nominal evidence for linkage to MS [2-6]. With the notable exception of the human leukocyte antigen (major histocompatibility complex [MHC]) locus on 6p21, evidence for specific MS genes emerging from these studies is still scanty. Thus far, the only associated non-HLA genes replicated in multiple populations are the PRKCA gene [7] and the recently reported IL2RA and IL7R genes [8]. For MS, as for most complex traits, the loci derived from linkage scans have remained quite wide because of multiple uncertainties concerning the disease model in statistical analyses. To expedite the process of gene identification in these wide DNA regions, we need novel approaches to identify potentially involved pathways and to prioritize genes on identified loci for further sequencing efforts.Our idea is to turn to full genome functional studies for these goals. As illustrated in Figure 1, our approach takes advantage of the availability of abundant microarray data and a wealth of genomic/proteomic knowledge base from the public domain. Our intention is to integrate information from both the candidate gene and the full genome scan (thus far mostly family-based linkage) approaches. In this report we use two previously reported MS susceptibility genes, identified in the same study sample [7,9], n
%U http://genomebiology.com/2007/8/10/R205