Erratum: The Human Transcription Factors (Cell (2018) 172(4) (650–665), (S0092867418301065) (10.1016/j.cell.2018.01.029))

Abstract

(Cell 172, 650–665; February 8, 2018) Our Review described recent progress in characterizing binding sites for human transcription factors and efforts to characterize them functionally. We have identified five minor errors in the Review. In the Results, the sentence reading “Among the 1,107 proteins with a known motif, less than 2% (19) lack a canonical DBD, with only 6 of 69 such proteins having an in vitro derived motif—the other 13 are based on experiments such as ChIP-seq and thus may describe binding through a cofactor” created ambiguity about how the different subsets of proteins were classified. To clarify this point, the sentence has been revised to read “Among the 1,107 TFs with a known motif, less than 2% (19) are among the 69 putative TFs that lack a canonical DBD. Indeed, among the 69, it appears that only a small proportion have been analyzed in detail for sequence specificity, while the inclusion of the rest of the 69 proteins among TFs is based on other lines of evidence such as known individual genomic target sites, or existence of a DNA-protein co-structure. Only 6 of the 19 motifs for the 69 noncanonical TFs have been derived in vitro, while the other 13 motifs are based on experiments such as ChIP-seq, and thus could potentially describe binding through a cofactor.” In Figure 2B, the vertical axis originally read “Number of within-family clusters.” For clarity, this has been revised to “Number of distinct motifs.” Information about how the number of distinct motifs were obtained is now contained in the Figure 2 legend which read “Motif diversity within each family, as measured by the number of clusters supported by the optimal silhouette value (Lovmar et al., 2005),” and has now been revised to “Motif diversity within each family. The number of within-family motif clusters (distinct motifs) was calculated as the number of clusters supported by the optimal silhouette value (Lovmar et al., 2005).” In Table 1, library construction for the Spec-Seq method was described incorrectly. The original text “Single step SELEX with a microarray synthesized library. The lower complexity library is useful for quantitatively measuring effects of binding site mutations using sequencing” has been revised to read “Single step SELEX using a synthesized library of degenerate sequences of interest. The lower complexity library is useful for quantitatively measuring effects of binding site mutations using sequencing.” Additionally, the column heading “Capability of Motif Discovery (approx. length in base pairs)” is incorrect as there can be gaps in the sequence. The heading has been revised to “Capability of de novo motif discovery (approx. length in base pairs with high information content).” These errors have now been corrected in the online version of the paper. We apologize for any inconvenience they may have caused to the readers. [Figure presented]