Fig. 6

Integrated chromatin state analysis in pre-B cells identifies age-related changes. a Emission parameters learnt by the Hidden Markov Model in chromHMM analysis. It shows the likelihood of emitting each of the 5 marks used in this analysis at any given chromatin state. b Transition parameters learnt by the HMM in chromHMM analysis: given the current chromatin state (at a specific 200-bp genomic region), how likely is it that the adjacent 200-bp region will be in any of the 16 inferred states. Dashed blocks show states that transit to each other more frequently suggesting a more biologically meaningful chromatin state, thus driving our clustering into 6 states: Polycomb, Bivalent, Active Promoter, Active Regulatory Region, Insulator, and Background. c Top: Heatmap showing the change in the occupancy (bp) of each regulatory chromatin state over gene promoter regions (2500-bp up- and downstream of TSS), for a subset of genes showing the largest magnitude changes. Genes are hierarchically clustered based on their correlation and assigned to 9 clusters showing distinct patterns. Bottom: Log2 fold change in expression of genes in each cluster, in aged versus young pre-B cells. Dashed line indicates unchanged gene expression. To better display the data, some outliers are not displayed. d Fraction of promoter regions (defined as for c) of DEGs occupied by each of the 6 chromatin states in young and aged pre-B cells. Note that Let-7b/c2 refers to the potential regulatory region/alternative promoter over which we have observed significant alterations in chromatin marks