Summary Recent breakthroughs in single-cell genomics allow probing molecular states of cells with unprecedented detail along the sequence of the DNA. Biological function relies, however, on emergent processes in the three-dimensional space of the nucleus, such as droplet formation through phase separation. Here, we use single-cell multi-omics sequencing to develop a theoretical framework to rigorously map epigenome profiling along the DNA sequence onto a description of the emergent spatial dynamics in the nucleus. Drawing on scNMT-seq multi-omics sequencing in vitro and in vivo we exemplify our approach in the context of exit from pluripotency and global de novo methylation of the genome. We show how DNA methylation patterns of the embryonic genome are established through the interplay between spatially correlated DNA methylation and topological changes to the DNA. This feedback leads to the predicted formation of 30-40nm sized condensates of methylated DNA and determines genome-scale DNA methylation rates. We verify these findings with orthogonal single cell multi-omics data that combine the methylome with HiC measurements. Notably, this scale of chromatin organization has recently been described by super-resolution microscopy. Using this framework, we identify local methylation correlations in gene bodies that precede transcriptional changes at the exit from pluripotency. Our work provides a general framework of how mechanistic insights into emergent processes underlying cell fate decisions can be gained by the combination of single-cell multi-omics and methods from theoretical physics that have not been applied in the context of genomics before. Highlights We develop methodology to infer collective spatio-temporal processes in the physical space of the nucleus from single-cell methylome sequencing experiments. We show that DNA methylation relies on a feedback between de novo methylation and nanoscale changes in DNA topology, leading to the formation of methylation condensates. Chromatin condensates at this scale have recently been described by high-resolution microscopy but have remained without mechanistic explanation. Using this framework, we identify changes in the distribution of DNA methylation marks in gene bodies that precede gene silencing at the exit from pluripotency.

Inference of emergent spatio-temporal processes from single-cell sequencing reveals feedback between<i>de novo</i>DNA methylation and chromatin condensates

Inference of emergent spatio-temporal processes from single-cell sequencing reveals feedback betweende novoDNA methylation and chromatin condensates