Wide-Scale Analysis of Human Functional Transcription Factor Binding Reveals a Strong Bias towards the Transcription Start Site

We introduce a novel method to screen the promoters of a set of genes with shared biological function, against a precompiled library of motifs, and find those motifs which are statistically over-represented in the gene set. The gene sets were obtained from the functional Gene Ontology (GO) classification; for each set and motif we optimized the sequence similarity score threshold, independently for every location window (measured with respect to the TSS), taking into account the location dependent nucleotide heterogeneity along the promoters of the target genes. We performed a high throughput analysis, searching the promoters (from 200bp downstream to 1000bp upstream the TSS), of more than 8000 human and 23,000 mouse genes, for 134 functional Gene Ontology classes and for 412 known DNA motifs. When combined with binding site and location conservation between human and mouse, the method identifies with high probability functional binding sites that regulate groups of biologically related genes. We found many location-sensitive functional binding events and showed that they clustered close to the TSS. Our method and findings were put to several experimental tests. By allowing a “flexible” threshold and combining our functional class and location specific search method with conservation between human and mouse, we are able to identify reliably functional TF binding sites. This is an essential step towards constructing regulatory networks and elucidating the design principles that govern transcriptional regulation of expression. The promoter region proximal to the TSS appears to be of central importance for regulation of transcription in human and mouse, just as it is in bacteria and yeast.

💡 Research Summary

The authors present a genome‑wide computational framework to identify functional transcription‑factor (TF) binding sites (BS) that are enriched in groups of genes sharing a biological function, and to assess how the distance of a BS from the transcription‑start site (TSS) influences its functionality. Gene sets were defined using 134 Gene Ontology (GO) categories, covering 8,110 human and 23,400 mouse genes. A library of 412 position‑specific scoring matrices (PSSMs) representing known TF motifs was compiled. For each GO‑motif pair the authors scanned promoter regions from –1000 bp upstream to +200 bp downstream of the TSS, dividing them into short windows (≈100–200 bp). Within each window a score threshold (T) was optimized separately for that GO set, thereby accounting for local nucleotide composition biases. Hits (sequences scoring ≥ T) were counted, and a hypergeometric test compared the observed hit count in the GO set to the expected count in a reference set consisting of all promoters. False‑discovery‑rate (FDR) control was applied to obtain statistically significant enrichments.

To further reduce false positives, the same analysis was repeated on orthologous mouse genes; only motif‑window enrichments that were reproduced in both species were retained, providing an evolutionary conservation filter that does not rely on strict sequence conservation across many species.

The resulting landscape revealed two major classes of GO‑motif combinations. One class displayed a pronounced positional bias: functional BSs were overwhelmingly concentrated in a ~300 bp region spanning –200 bp to +100 bp relative to the TSS. Motifs such as the TATA box, NF‑κB, E2F, Myc, and several cell‑cycle regulators fell into this group. In contrast, the same motifs located further upstream (‑600 bp to ‑1000 bp) showed no significant enrichment, indicating that distance from the TSS is a critical determinant of functional impact for many TFs.

Experimental validation was performed in two ways. First, the authors cross‑referenced their predictions with published ChIP‑seq datasets, confirming that a large fraction of the predicted BSs are indeed bound in vivo. Second, reporter‑gene assays were conducted in which a TF binding site was placed either near the TSS or at a distal upstream position; transcriptional activation was dramatically stronger when the site resided within the proximal window, corroborating the computational inference.

The study demonstrates that (i) a location‑specific enrichment analysis, coupled with GO‑based functional grouping, can uncover TF‑BS relationships that are missed by conventional methods relying solely on sequence conservation; (ii) the proximal promoter region (≈‑200 bp to +100 bp) plays a central, perhaps universal, role in transcriptional regulation in mammals, analogous to the well‑established situation in bacteria and yeast.

Limitations include reliance on GO annotations (which may not perfectly reflect co‑regulation), the use of PSSM scores that ignore DNA shape, nucleosome positioning, and epigenetic modifications, and the binary conservation filter that only considers human–mouse orthology. The authors suggest that integrating ATAC‑seq, DNase‑seq, Hi‑C, and deep‑learning‑based binding predictors could further refine the approach.

In summary, this work provides a robust, statistically grounded method for detecting functional TF binding sites, reveals a strong bias of functional sites toward the transcription start site, and offers a valuable resource for constructing accurate regulatory networks in higher eukaryotes.