In this R21 application we propose to implement improvements to direct genomic selection to more rapidly capture collections of highly conserved sequence elements across diverse genomes and diverse species. We shall also implement improvements that are more targeted towards the non-conserved sequences that adjoin conserved elements, in order to capture a more complete coverage of orthologous regions between individuals and species. These methodological improvements should have broad relevance and interest to many areas of human genetics and comparative genomics. ? ? To increase the speed in which direct selection experiments can be performed. The most ? time consuming aspect of the current selection method is the length of hybridization time required to reach a Cot1/2 of 25 to 100. Most investigators would prefer this to be simplified and standardized into a """"""""kit-like"""""""" format. We shall test phenol emulsion methods to increase nucleic acid reassociation rates and to standardize the selection parameters. To explore methods for enhancing the coverage of direct selection. We shall extend the sequence coverage of direct selection by standardizing random shearing methods to generate larger fragment lengths. We shall employ these, along with long-range PCR methods, to capture more of the non-conserved or repetitive sequences across large contigs. We shall also construct and use biologically amplifiable, singlestranded libraries of genomic DNAs from which coverage can be extended in a secondary selection step. To increase the target genomic interval for selections to greater than 1Mb. Many largescale positional cloning projects, whether for """"""""simple"""""""" or complex diseases, result in the identification of large genomic target intervals (~1Mb). Likewise, many cross-species comparative genomics projects would benefit from the ability to target large regions. Our preliminary data indicate that 470kb can be efficiently employed as a target for direct selection. We shall demonstrate proof of principle that >1Mb can be efficiently used in these methods. To demonstrate the feasibility of conducting cross-species direct selections on a ~1Mb interval in 20 evolutionarily diverged species. Comparative genomics is critically dependent on obtaining large ? scale genomic DNA sequence information across diverse species. We shall apply selection methods to a genomic region that has biological relevance to craniofacial morphogenesis in many species (including humans). We shall employ targeted selections (and a limited amount of comparative DNA sequencing) across the Bmp4 locus in avian species that had last common ancestors between 40 million and 120 million years ago. ? ? ?
