Submitted to: Plant and Animal Genome Conference
Publication Type: Proceedings
Publication Acceptance Date: November 1, 2009
Publication Date: January 9, 2010
Citation: Liu, G., Cellamare, A., Ventura, M., Li, C., Eichler, E.E. 2010. Comparison of the segmental duplication pattern on two cattle genome assemblies using FISH. Plant and Animal Genome Conference. W178.
We previously estimated that 3.1% (94.4 Mb) of the bovine genome consists of recently duplicated sequences (>= 1kb in length, >= 90% sequence identity) using two distinct computational analyses (WGAC and WSSD). In this study, we further validated selected large duplications and compared their distribution on two independent cattle genome assemblies (Btau_4.0 and UMD3) using fluorescent in situ hybridization (FISH). Since our BAC-FISH method was most reliable to distinguish single signal vs. duplicated signals and interchromosomal vs. intrachromosomal duplications, our assembly comparisons were based on these two criteria. A total of 46 BAC clones were selected from high-confidence duplicated regions (Both, WSSD and WGAC<94%) and over 80% (33/41) of them produced multiple signals by examination of interphase and metaphase FISH confirming their duplication status. Furthermore, the two assemblies essentially produced the same computational duplication predictions for those 46 BAC clones, suggesting they are almost identical in those high-confidence duplicated regions. On the other hand, as assembly artifacts are particularly enriched in large, high-identity intrachromosomal duplications identified only by WGAC (which is assembly-dependent), we chose 13 BAC clones from those low-confidence duplicated regions as FISH probes to compare Btau_4.0 and UMD3. Within 12 BAC clones which produced FISH results, 10 BAC clones supported UMD3’s predictions while only 2 supported Btau_4.0’s. This result is a striking contrast to the result of the above 46 BAC clones, suggesting that majority of the large, high-identity duplications (WGAC+ only) in Btau_4.0 are probably assembly artifacts. In the light of neither of these two assemblies is perfect in terms of totally agreeing with the FISH results, our results demonstrate that segmental duplications have been problematic for genome sequencing and assembly.