Transcription factors (TFs) bind DNA cis-regulatory elements (CREs) in the context of chromatin to regulate the transcription machinery and gene expression. The precise positioning of the pre-initiation complex (PIC), TFs, cofactors, and nucleosomes within the human genome—and their interactions on a genomic scale—has yet to be thoroughly investigated. In this study, we developed an advanced version of the ChIP-exo assay (v6), utilizing a combination of nucleases as high-precision structural probes to reconstruct these interactions with single-base genomic resolution. We show that for many TFs, their unbound CREs are typically translationally and/or rotationally buried on a nucleosome. For CTCF/Cohesin binding, not only do nucleosomes become translationally phased at its flanks, but CTCF/Cohesin retains interactions with each of them. For pioneer factor FoxA binding, the sites generally lack a rotational and translational setting on nucleosome. For NFIA binding, the CRE becomes rotationally exposed and translationally constrained at the nucleosome edge. Nuclease probes also reveal precise and general architectural relationships among TFs (SP1, GABPA), cofactors (P300), the transcription machinery (TBP, TFIIB, Pol II), and nucleosomes at promoters. We propose a refined model of human promoters characterized by bidirectional transcription, where two distinct PICs initiate on opposing strands while sharing common upstream regulatory regions. Our findings revealed that different general transcription factors (GTFs) exhibit specific ChIP exonuclease stop patterns at TATA boxes, which are precisely located 30 bp upstream of genome-wide transcription start sites (TSSs), irrespective of strand orientation. Furthermore, when RNA polymerase II (Pol II) transitions into a transcriptionally paused state, TBP and TFIIA show significantly higher occupancy at TATA-like core promoters compared to TATA-less core promoters. Additionally, we discovered that genome-wide +1 nucleosomes exhibit a stable in vivo architecture, with paused Pol II embedded within these rotationally constrained nucleosomes. The rotational setting of the +1 nucleosome is influenced primarily by the periodic occurrence of specific dinucleotides but is also regulated by cofactors. Notably, we identified a novel motif that is both translationally and rotationally phased on the +1 nucleosome. When this motif is bound by WDR5/RbBP5, the DNA is rotated by 5 bp (or 10n + 5 bp), leading to an alteration in its rotational phasing.