Fig. 1: Strategy to map the full compendium of human nuclear receptor-binding sites.
From: Hidden modes of DNA binding by human nuclear receptors
a A MinSeq set is extracted from the DNA–protein interactome (DPI) of full-length human nuclear receptors (NRs) using the MinSeq Find algorithm. The Peak Assign analysis uses the extracted MinSeq set to score and annotate chromatin immunoprecipitation (ChIP-seq)–derived genomic loci bound by NRs in vivo. In parallel, the SNP Align analysis evaluates the impact of single nucleotide polymorphisms (SNPs) on creating or disrupting NR-binding sites. b Phylogenetic tree (neighbor-end joining) of NRs and corresponding small-molecule ligands (numbered in green boxes) used in this study. (Source data are provided as a Source Data file). c A schema of the MinSeq Find algorithm. DNA-sequencing reads obtained via high throughput-SELEX are counted as patterns of nucleotide sequences of different lengths separated by linkers of varying sizes, referred to here as MinSeq (Online Methods). MinSeq Find can capture multiple patterns of binding, including a variable spacer sequence and length and NR-binding orientation. Fold enrichment for MinSeqs is calculated by normalization of the read count against a PAGLO library model. The iterative algorithm Orthogonal Matching Pursuit (OMP) then further minimizes and optimizes the MinSeq set. (Designed by Laura Vanderploeg).
