pool and a reduced cortical neuron number (24).
Heightened signaling through FGFR1 during rat
corticogenesis increases the neuron number by
more than 80% (25).
Module 15 is enriched in human H3K27ac and
H3K4me2 promoter gains [Benjamini-Hochberg
(BH)–corrected permutation, P = 0.003] (fig.
S10A) associated with cortical-patterning ontologies such as regionalization (binomial test, P =
7.53 × 10−4) and forebrain development (
binomial test, P = 1.63 × 10−5) (fig. S10B). Homeobox
genes are notably enriched among genes associated with gains in this module (binomial test, P =
1.18 × 10−8).
Module 10 shows the strongest enrichment of
human lineage H3K27ac and H3K4me2 promoter
and enhancer gains in the network (Fig. 3, A and
B). Genes implicated in extracellular matrix (ECM)
functions are significantly overrepresented among
gain-associated genes in this module (binomial
test, P = 2.26 × 10−7) (table S5). The ECM contributes to the maintenance of human progenitor cell self-renewal and neuronal migration (26).
Module 10 gain-associated genes are also enriched
Fig. 3. Enrichment of
epigenetic gains in
module 10. (A) Epigenetic
gains mapped onto
module 10; genes
associated with gains are
highlighted as in Fig. 2B.
(B). Fold enrichment of
H3K27ac promoter or
enhancer gains at each
human time point in this
module. *P < 0.01
(BH-corrected permutation). (C) Genes in the
related FGF, TGFb, bone
(BMP), and ECM signaling
pathways are associated
with gains from module 10
(yellow stars) and module
3 (red stars). Genes or
gene families are highlighted in orange; associated biological
processes are in green.
The pathway shown is
derived from KEGG
Fig. 4. Modules enriched for epigenetic gains converge on common biological processes. (A) Gain-enriched modules exhibit significantly higher gene expression correlation values with each other than with
modules not enriched for gains. *P < 1.1 × 10−15 (Wilcoxon rank sum test). (B) Eigengene expression
correlations among the top 35 modules (as ranked by number of genes). Modules enriched in gains are
numbered. Ontologies associated with gains in each module are highlighted. Arrows connect modules that
include each transcription factor shown with modules that are enriched for that factor’s binding motif.