Centromeric variation has been linked to cancer and infertility, but centromere sequences contain multiple tandem repeats and can only be assembled manually from long error-prone reads. Here we describe the centroFlye algorithm for centromere assembly using long error-prone reads, and apply it to assemble human centromeres on chromosomes 6 and X. Our analyses reveal putative breakpoints in the manual reconstruction of the human X centromere, demonstrate that human X chromosome is partitioned into repeat subfamilies and provide initial insights into centromere evolution. We anticipate that centroFlye could be applied to automatically close remaining multimegabase gaps in the reference human genome.

着丝粒变异与癌症和不孕不育有关，但着丝粒序列包含多个串联重复序列，只能从容易出错的长读数中手动组装。在这里，我们描述了使用容易出错的长读段进行着丝粒组装的 centroFlye 算法，并将其应用于在 6 号和 X 号染色体上组装人类着丝粒。我们的分析揭示了人类 X 着丝粒手动重建中的假定断点，证明人类 X 染色体是分为重复亚家族并提供对着丝粒进化的初步见解。我们预计 centroFlye 可用于自动闭合参考人类基因组中剩余的多兆碱基间隙。