Here we present a novel method: single-cell combinatorial indexing for methylation analysis (sci-MET), which is the first highly scalable assay for whole genome methylation profiling of single cells. We use sci-MET to produce 2,697 total single-cell bisulfite sequencing libraries and achieve read alignment rates of 69 ± 7%, comparable to those of bulk cell methods. As a proof of concept, we applied sci-MET to successfully deconvolve the cellular identity of a mixture of three human cell lines.. CC-BY-NC-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint . http://dx.doi.org/10.1101/157230 doi: bioRxiv preprint first posted online Jun. 28, 2017;
MainThe fundamental challenge of identifying and characterizing the molecular properties of every cell type in the body has recently entered the realm of possibility 1 . High-throughput singlecell transcriptome [2][3][4] and chromatin accessibility profiling assays [5][6][7] have dramatically improved our ability to uncover latent cell types within complex tissues and dynamic cell states during differentiation. These same possibilities extend to DNA methylation; a covalent addition of a methyl group to cytosine bases in the genome that largely serves in a repressive role 8 . DNA methylation occurs at a high rate in cytosine-guanine dinucleotides (CG), with cytosine methylation in non-CG sites (CH) occurring rarely and only in select tissues 9 . Both CG and CH methylation have cell type-specificity and are the subject of active modification in developing tissues 8 . DNA methylation can be probed at base-pair resolution using the deaminating chemistry of sodium bisulfite treatment before or after the generation of sequencing libraries (e.g. as in whole genome bisulfite sequencing, WGBS) 10,11 . Despite the comprehensive nature of these assays, key aspects of methylation architecture and dynamics remain elusive, with cell type heterogeneity at the forefront of this challenge.Recent work has optimized bisulfite sequencing to decrease input requirements to the single cell level (scWGBS) [12][13][14][15] . These assays have provided unique insights into environmental effects on methylome dynamics 13 , have uncovered new cell-types that form during hematopoesis 14 , and been coupled with scRNA-seq to directly study the relationship of DNA methylation to transcription 15 . However, these methods use a parallelized library generation strategy in which the WGBS protocol must be carried out on each cell in its own reaction vessel and thus remain low-throughput. Existing platforms to assess the transcriptome of single cells in high-throughput largely rely on droplet-based microfluidics strategies by sequestering single cells into individual droplets along with a substrate harboring barcoded oligonucleotides for cell-level indexing 2,16 . This approach is not readily adaptable to genome-wide DNA methylation pro...