Extended Data Figure 3: Relationship between transcription-factor binding motifs and occupancy, and LSI clade-specific accessibility.

a–c, SeqGL was run on LSI clade-specific distal peaks at each time point to identify enriched sequence motifs. The top five most-enriched unique motifs for each clade are displayed. Coloured circles indicate the clade represented by each line. For the later time points (6–8 h and 10–12 h), blue is neurogenic ectoderm, yellow is non-neurogenic ectoderm, red is myogenic mesoderm and green is mesendoderm. The results show an enrichment of motifs for factors associated with early development at 2–4 h with more tissue-specific factor motifs (for example, mesodermal factor Mef2 or neural regulator Tramtrack) within germ-layer annotated clades at later stages of development. d–l, Using ChIP occupancy data (peaks) and transcription factor binding motifs compiled previously¹⁶, we scanned for all transcription factor motif instances under ChIP peaks from datasets spanning 6–8 h of development using FIMO. Aggregate read counts in 4-kb windows centred on each identified motif instance are shown for each of the four LSI clades at 6–8 h. Green, endoderm; red, myogenic mesoderm; yellow, non-neurogenic ectoderm; blue, neurogenic ectoderm. Light shading in the same colours indicates 95% confidence intervals. d–g, Aggregate plots for four ubiquitous transcription factors (BEAF32, CTCF, Pho, and Trl) at 6–8 h. h–l, Aggregate plots for mesodermal transcription factors (Bap, Lmd, Mef2, Tin, Twi) at 6–8 h.