文章题为：An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues

概述

William J Greenleaf和Howard Y Chang继开发ATAC-seq之后又一升级版ATAC-seq protocol

omni-ATAC, an improved ATAC-seq protocol for chromatin accessibility profiling that works across multiple applications with substantial improvement of signal-tobackground ratio and information content.

特殊的细胞类型和组织需要针对性地对protocol进行优化，这使得这类数据难以在不同研究中进行比较。因此作者报道了一种提升的、可广泛应用的ATAC-seq protocol，叫做Omni-ATAC，对于多种细胞系、组织类型、长期保存的冷冻样本都可适用且同时能提升数据的质量

提升

protocol的优化主要体现在三个方面：

1. 多种去污剂的使用：比如NP40, Tween-20, and digitonin。以提高不同细胞类型的通透性并用于在转座反应中去除线粒体
2. 裂解后使用Tween-20洗涤的步骤，进一步去除线粒体以及增加文库复杂度
3. 在转座反应中使用PBS增强信噪比

与先前的ATAC-seq相比，Omni-ATAC protocol大大降低了测序成本，产生的mapping到线粒体上的reads降低到前者的1/13，并且提高了数据质量，比原来多三倍的reads落在peaks中（也就是FRiP值增加），每个细胞的unique fragments是原来的15倍。

compare_QC.jpg

上图比对了Omni-ATAC和标准ATAC、Fast-ATAC的质控信息，包括文库大小、线粒体reads百分比、TSS富集信号

另一方面Omni-ATAC的灵敏度也得到了大幅提升，可以检测出许多标准ATAC检测不到的peaks。

其他优点还包括进一步降低input细胞数、可应用于快速冷冻存储的样本（因为原来的standard ATAC-seq and Fast-ATAC都要求转座反应的样本必须是fresh sample)

验证

于是作者测试了一下Omni-ATAC处理临床冷冻样本（比如脑组织样本）的效果：

clinical_performance.jpg

作图展示的是样本之间基于peaks的无监督聚类结果，可以看到相同脑组织区域的样本聚在一起

右图示结合了GWAS的SNP mapping结果，典型的如海马体（与记忆、遗忘有关）恰好其开放染色质区富集了阿尔兹海默症的SNP

除此之外，作者还应用了免疫组化等技术佐证了其Omni-ATAC的可靠性：

histology.jpg

通过对ATAC-seq所取组织的邻近组织进行染色验证

补充

作者在supplementary材料中补充说明了他们这样优化protocol的原理和目的：

先是列举了详细的protocol优化细节：

Rationale and Efficacy of Chosen Optimizations

The individual optimizations made to the original ATAC-seq protocol that comprise the Omni-ATAC protocol are as follows:
(1) The addition of 0.1% Tween-20 and 0.01% digitonin to the NP40-based lysis buffer
(2) Washing lysed cells in a resuspension buffer containing 0.1% Tween-20
(3) The addition of PBS, 0.1% Tween-20, and 0.01% digitonin to the transposition mix

然后说明原因：

The addition of Tween-20 and digitonin to the NP40-based lysis buffer

(i) leads to a reduction in mitochondrial reads and an increase in library complexity and (ii) enables lysis in a wider array of cell types. Washing of lysed cells in resuspension buffer containing Tween-20 on its own does not improve data quality, but in the context of lysis with Tween-20, digitonin, and NP40, further increases the library complexity and further reduces reads from mitochondrial DNA.

The addition of PBS to the transposition reaction on its own increases the enrichment of signal at TSSs by 48% on average.

Collectively, these optimizations (i) improve ATAC-seq data quality in cell types and cell contexts that were already tractable with the standard ATAC-seq protocol and (ii) make ATAC-seq possible in cell types and cell contexts that were not possible with the standard ATAC-seq protocol.

其他补充还有适当降低Tn5转座酶的浓度等。

关于文中提到的fast-ATAC，可以参考这篇：https://doi.org/10.1038/ng.3646

fast_atac.jpg

实际上，ATAC-seq系列仍然是有一些不足之处的：

在样本的离心等机械处理过程中，有些本来是结合的染色质可能变得开放并被转座酶切割产生假阳性；

仅有一半的分子加上的序列接头是能PCR扩增的（可能有的片段接上的两个接头一样）

加接头的位点之间的距离可能不适合PCR扩增反应（PCR扩增有长度限制）

参考：http://enseqlopedia.com/wiki-entry/dna-sequencing-methods/dna-protein-interactions/atac-seqfast-atac/

参考文献

An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues：https://www.nature.com/articles/nmeth.4396

Lineage-specific and single-cell chromatin accessibility charts human hematopoiesis and leukemia evolution：https://www.nature.com/articles/ng.3646?proof=true#citeas