Comparative genomics revealed that ~5% of the human genome is conserved among mammals. This fraction is likely functional, and could harbor pathogenic mutations. We have shown (Nature 2002, Science 2003) that more than half of the constrained fraction of the genome consists of Conserved Non-Coding sequences (CNCs). Model organisms provided evidence for enhancer activity for a fraction of CNCs; in addition another fraction is part of large non-coding RNAs (lincRNA). However, the function of the majority of CNCs is unknown. Importantly, a few pathogenic mutations in CNCs have been associated with genetic disorders. We propose to i) perform functional analysis of CNCs, and ii) identify the spectrum of pathogenic CNC mutations in recognizable human phenotypes. The aims are: 1. Functional genomic connectivity of CNCs 1a. Use 4C in CNCs in various cell types and determine their physical genomic interactions. 1b. Perform targeted disruption of CNCs in cells and assess the functional outcomes. 2. Pathogenic variation of CNCs 2a. Assess the common variation in CNCs: i) common deletion/insertions in 350 samples by aCGH of all human CNCs; ii) common SNP/small indels using DNA selection and High Throughput Sequencing (HTS) of CNCs in 100 samples. 2b. Identify likely pathogenic mutations in developmental syndromes. Search for i) large deletions and duplications of CNCs using aCGH in 1500 samples with malformation syndromes, 1000 from spontaneous abortions, and 500 with X-linked mental retardation; and ii) point mutations in these samples by targeted HTS. The distinction between pathogenic and non-pathogenic variants is difficult, and we propose approaches to meet the challenge. 3. Genetic control (cis and trans eQTLs) of expression variation of CNC lincRNAs, using 200 samples.