Back

Beyond motif recognition: Specificity of human transcription factors in yeast

Bugis, J.; Zlotnik-Weinberg, D. R.; Manadre, W.; Mindel, V.; Lu, Y.; Fuxman Bass, J. I.; Barkai, N.

2026-04-29 genetics
10.64898/2026.04.27.721015 bioRxiv
Show abstract

Transcription factors (TFs) bind DNA through sequence-specific DNA-binding domains (DBDs), yet genome-wide analyses show that TFs occupy only a small fraction of their motif occurrences. This raises the question of how TFs distinguish specific targets from the many potential sites in the genome. To investigate determinants of binding specificity beyond the cognate motif and cofactor influences, we measured the binding of 60 human TFs across the budding yeast genome. Although human TFs robustly recognized their motifs, they displayed strong selectivity in site occupancy. Nucleosome abundance explained this selectivity only in part: among the 5-20% of motif sites that were bound, a substantial fraction remained nucleosome covered. Furthermore, TFs recognizing similar motif sequences independently localized to distinct subsets of sites within different promoters. Despite the absence of human-specific cofactors in yeast, both binding stability and genomic preferences depended on largely disordered non-DBD regions. These findings suggest intrinsically disordered regions (IDRs) may therefore direct genome binding TF target recognition across evolutionarily distant genomes. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=124 SRC="FIGDIR/small/721015v1_ufig1.gif" ALT="Figure 1"> View larger version (31K): org.highwire.dtl.DTLVardef@7db4cborg.highwire.dtl.DTLVardef@e88903org.highwire.dtl.DTLVardef@7b3e11org.highwire.dtl.DTLVardef@127ef29_HPS_FORMAT_FIGEXP M_FIG C_FIG

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.