Fig. 1

Overview of the SPIN method and the SPIN states in K562. a Left panel: Input data for SPIN, including 1D genomic signals (TSA-seq and DamID), and Hi-C data. Middle panel: Graph representation of the hidden Markov random field. Genomic bins are represented as nodes and the hidden states of nodes are labeled by different colors. Right panel: Cartoon illustration of the SPIN states output for the nodes. b Genome browser view of the input data (TSA-seq and DamID) and output (SPIN states). Three regions with different SPIN states (Interior_Act3, Lamina, and Speckle) are highlighted. These regions have distinct distributions of TSA-seq and DamID signals. c Boxplots showing the distributions of the input TSA-seq and DamID data on different SPIN states. The genome-wide percentage (size) for each SPIN state is also shown. d Cartoon of the spatial positions of SPIN states relative to different nuclear compartments (nuclear lamina, nuclear speckles, nucleoli). e Comparison between the identified SPIN states and DNA FISH. Three genomic regions with different SPIN states (Lamina chr18:41,032,946-41,237,107, Interior_Act1 chr18:48,801,892-48,998,008, and Speckle chr7:100,470,711-100,665,335) are shown. DNA probes are labeled in red (with white arrowhead) and nuclear speckles (SON protein) is labeled in green. Nuclear DNA staining (DAPI) is in blue