[½»Á÷] ÇëÎʸ÷λ´óÅ£ÃÇÖªµÀÈ«»ùÒò×éthree-way alignments Ôõô²Ù×÷µÄÂð

RT£¬ÇëÎʸ÷λ´óÅ£ÃÇÖªµÀÈ«»ùÒò×éthree-way (species) whole-genome sequence alignments in MAF format (UCSC)Ôõô²Ù×÷µÄÂð To perform an exhaustive, genome-wide screening for phyloge- netically informative retroposon insertions in placentals and to test the three viable phylogenetic hypotheses of the basal origin of placentals, we derived two different three-way (species) whole-genome sequence alignments in MAF format (UCSC): (1) armadillo¨Celephant¨Chuman (3050 gigabases) and (2) elephant¨C human¨Carmadillo (2730 gigabases), the decisive difference between the two alignments being the full retroposon represen- tation of the leading species that was used as the profile for the remaining species. The three-way alignments were prepared using pairwise (armadillo¨Celephant and armadillo¨Chuman for the first three-way alignment and elephant¨Chuman and armadillo¨Celephant for the second three-way alignment) BLASTZ alignments (Chiaromonte et al. 2002; Kent et al. 2003; Schwartz et al. 2003) and the standard matrix used in the UCSC browser. The alignments were fed into axtChain with the parameters chainMinScore = 3000  chain- LinearGap = medium, which organizes all alignments by chro- mosome (or scaffold) and creates a kd-tree from the gapless subsections (blocks) of the alignments. A dynamic program was then run over the kd-trees to find themaximally scoring chains of these blocks. Finally, MULTIZ (Blanchette et al. 2004) was used to generate the two three-way alignments referencing elephant and armadillo. We used the first alignment to search for elements supporting the Atlantogenata [(armadillo + elephant)  human] and Exa- froplacentalia [(armadillo + human)  elephant] hypotheses, and the second one to test the Epitheria [(elephant + human)  armadillo] hypothesis. ÕâÊÇÔ­ÎÄ£¬ÎÒ¶ÔÉúÎïÐÅÏ¢Ò»µã»ù´¡¶¼Ã»ÓУ¬ÎÞÄÎÀÏ°å·ÇÒªÎÒ×ö£¬ÇóÖú°¡

»Ø¸´´ËÂ¥