Parida Laxmi, Melé Marta, Calafell Francesc, Bertranpetit Jaume
Computational Biology Center, IBM TJ Watson Research, Yorktown Heights, NY 10598, USA.
J Comput Biol. 2008 Nov;15(9):1133-54. doi: 10.1089/cmb.2008.0065.
Traditionally nonrecombinant genome, i.e., mtDNA or Y chromosome, has been used for phylogeography, notably for ease of analysis. The topology of the phylogeny structure in this case is an acyclic graph, which is often a tree, is easy to comprehend and is somewhat easy to infer. However, recombination is an undeniable genetic fact for most part of the genome. Driven by the need for a more complete analysis, we address the problem of estimating the ancestral recombination graph (ARG) from a collection of extant sequences. We exploit the coherence that is observed in the human haplotypes as patterns and present a network model of patterns to reconstruct the ARG. We test our model on simulations that closely mimic the observed haplotypes and observe promising results.
传统上，非重组基因组，即线粒体DNA（mtDNA）或Y染色体，已被用于系统地理学研究，特别是因其便于分析。在这种情况下，系统发育结构的拓扑是一个无环图，通常是一棵树，易于理解且在一定程度上易于推断。然而，对于基因组的大部分而言，重组是一个不可否认的遗传事实。出于更全面分析的需要，我们着手解决从一组现存序列估计祖先重组图（ARG）的问题。我们利用在人类单倍型中观察到的连贯性作为模式，并提出一种模式网络模型来重建ARG。我们在紧密模拟观察到的单倍型的模拟实验上测试我们的模型，并观察到了有前景的结果。
