An expanded view of complex traits from polygenic to omnigenic

Leading Edge
Perspective
An Expanded View of Complex Traits:
From Polygenic to Omnigenic
Evan A. Boyle,1,* Yang I. Li,1,* and Jonathan K. Pritchard1,2,3,*
1Department of Genetics
2Department of Biology
3Howard Hughes Medical Institute
Stanford University, Stanford, CA 94305, USA
*Correspondence: eaboyle@stanford.edu (E.A.B.), yangili@stanford.edu (Y.I.L.), pritch@stanford.edu (J.K.P.)
http://dx.doi.org/10.1016/j.cell.2017.05.038
A central goal of genetics is to understand the links between genetic variation and disease. Intui-
tively, one might expect disease-causing variants to cluster into key pathways that drive disease
etiology. But for complex traits, association signals tend to be spread across most of the
genome—including near many genes without an obvious connection to disease. We propose
that gene regulatory networks are sufficiently interconnected such that all genes expressed in dis-
ease-relevant cells are liable to affect the functions of core disease-related genes and that most
heritability can be explained by effects on genes outside core pathways. We refer to this hypothesis
as an ‘‘omnigenic’’ model.
The longest-standing question in genetics is to understand how
genetic variation contributes to phenotypic variation. In the early
1900s, there was fierce debate between the Mendelians—who
were inspired by Mendel’s work on pea genetics and focused
on discrete, monogenic phenotypes—and the biometricians,
who were interested in the inheritance of continuous traits
such as height. The biometricians believed that Mendelian ge-
netics could not explain the continuous distribution of variation
observed for many traits in humans and other species.
This debate was resolved in a seminal 1918 paper by R.A.
Fisher, who showed that, if many genes affect a trait, then the
random sampling of alleles at each gene produces a
continuous, normally distributed phenotype in the population
(Fisher, 1918). As the number of genes grows very large, the
contribution of each gene becomes correspondingly smaller,
leading in the limit to Fisher’s famous ‘‘infinitesimal model’’
(Barton et al., 2016).
Despite the success of the infinitesimal model in describing
inheritance patterns, especially in plant and animal breeding,
it was unclear throughout the 20th century how many genes
would actually be important for driving complex traits. Indeed,
human geneticists expected that even complex traits would be
driven by a handful of moderate-effect loci—thus giving rise to
large numbers of mapping studies that were, in retrospect,
greatly underpowered. For example, an elegant 1999 analysis
of allele sharing in autistic siblings concluded from the lack of
significant hits that there must be ‘‘a large number of loci
(perhaps R15).’’ This prediction was strikingly high at the
time but seems quaintly low now (Risch et al., 1999; Weiner
et al., 2016).
Since around 2006, the advent of genome-wide association
studies, and more recently exome sequencing, has provided
the first detailed understanding of the genetic basis of complex
traits. One of the early surprises of the GWAS era was that, for
typical traits, even the most important loci in the genome have
small effect sizes and that, together, the significant hits only
explain a modest fraction of the predicted genetic variance.
This has been referred to as the mystery of the ‘‘missing herita-
bility’’ (Manolio et al., 2009). The mystery has since been largely
resolved by analyses showing that common single-nucleotide
polymorphisms (SNPs) with effect sizes well below genome-
wide statistical significance account for most of the ‘‘missing
heritability’’ of many traits (Yang et al., 2010; Shi et al., 2016).
Rare variants with larger effect sizes also contribute genetic vari-
ance (Marouli et al., 2017), especially for diseases with major
fitness consequences (Simons et al., 2014) such as autism and
schizophrenia (De Rubeis et al., 2014; Fromer et al., 2014; Purcell
et al., 2014).
A second surprise was that, in contrast to Mendelian dis-
eases—which are largely caused by protein-coding changes
(Botstein and Risch, 2003)—complex traits are mainly driven
by noncoding variants that presumably affect gene regulation
(Pickrell, 2014; Welter et al., 2014; Li et al., 2016). Indeed,
many studies have shown that significant variants are highly
enriched in regions of active chromatin such as promoters and
enhancers in relevant cell types. For example, risk variants for
autoimmune diseases show particular enrichment in active chro-
matin regions of immune cells (Maurano et al.; 2012; Farh et al.,
2015; Kundaje et al., 2015).
These observations are generally interpreted in a paradigm in
which complex disease is driven by an accumulation of weak
effects on the key genes and regulatory pathways that drive
disease risk (Furlong, 2013; Chakravarti and Turner, 2016).
This model has motivated many studies that aim to dissect
the functional impacts of individual disease-associated variants
(Smemo et al., 2014; Sekar et al., 2016) or to aggregate hits to
identify key disease pathways and processes (Califano et al.,
2012; Jostins et al., 2012; Wood et al., 2014; Krumm et al.,
Cell 169, June 15, 2017 ª 2017 Elsevier Inc. 1177

2015). For several diseases, the leading hits have indeed
helped to highlight specific molecular processes—for example,
uncovering the role of autophagy in Crohn’s disease (Jostins
et al., 2012), and roles for adipocyte thermogenesis (Claussnit-
zer et al., 2015) and central nervous system genes in obesity
(Locke et al., 2015).
But despite the success of these earlier studies, we argue that
the enrichment of signal in relevant genes is surprisingly weak
overall, suggesting that prevailing conceptual models for com-
plex diseases are incomplete. We highlight some pertinent fea-
tures of current data and discuss what these may tell us about
the genetic architecture of complex diseases.
Distribution of GWAS Signals across the Genome
Early practitioners of GWAS were dismayed to find that, for most
traits, the strongest genetic associations could explain only a
small fraction of the genetic variance (Manolio et al., 2009).
This was taken to imply that there must be many causal loci,
each with small effect sizes (Goldstein, 2009). Subsequent ana-
lyses soon provided direct evidence for this in the case of schizo-
phrenia (Purcell et al., 2009) and showed that, together, common
variants can explain much of the expected heritability (Yang
et al., 2010). While traits vary greatly in terms of both the impor-
tance of the largest-effect common variants and of higher-pene-
trance rare variants (Loh et al., 2015; Shi et al., 2016; Sullivan
et al., 2017), it is now clear that polygenic effects are important
across a wide variety of traits (Shi et al., 2016; Weiner
et al., 2016).
One key question that has been under-studied to date is the
extent to which causal variants are spread widely across the
genome or clumped into disease-relevant pathways. However,
it is known that the heritability contributed by each chromosome
tends to be closely proportional to its physical length (Visscher
et al., 2006; Shi et al., 2016), hinting that causal variants may
be fairly uniformly distributed. And recent data show that causal
variants can be surprisingly dispersed even at finer scales. A pa-
per from Alkes Price and colleagues estimated that 71%–100%
of 1-MB windows in the genome contribute to heritability for
schizophrenia (Loh et al., 2015).
Here we explore a second example—namely, height—for
which very large GWAS datasets are available (Figure 1). While
height is often thought of as the quintessential polygenic trait,
recent work shows that the genetic architecture of height is actu-
ally broadly similar to that of a wide variety of other quantitative
traits and diseases ranging from diabetes or autoimmune dis-
eases to BMI or cholesterol levels. Thus, we use height to illus-
trate the extreme polygenicity typical of many complex traits
(Shi et al., 2016; Chakravarti and Turner, 2016).
A height meta-analysis from the GIANT study reported 697
genome-wide significant loci that, together, explain 16% of
the phenotypic variance (Wood et al., 2014). But a quantile-
quantile plot comparing the distribution of p values against
the expected null distribution shows that the distribution of p
values is hugely shifted toward small p values (Figure 1A),
such that common variants together explain 86% of the ex-
pected heritability (Shi et al., 2016). The inflation is stronger
in active chromatin and in expression quantitative trait loci
(eQTLs), consistent with the expected enrichment of signal
in gene-regulatory regions.
We next used ashR to analyze the distribution of regression
coefficients from the set of all SNPs (Stephens, 2017). ashR
models the GWAS results as a mixture of SNPs that have a
Figure 1. Genome-wide Signals of Association with Height
(A) Genome-wide inflation of small p values from the GWAS for height, with particular enrichment among expression quantitative trait loci and single-nucleotide
polymorphisms (SNPs) in active chromatin (H3K27ac).
(B) Estimated fraction of SNPs associated with non-zero effects on height (Stephens, 2017) as a function of linkage disequilibrium score (i.e., the effective number
of SNPs tagged by each SNP; Bulik-Sullivan et al., 2015b). Each dot represents a bin of 1% of all SNPs, sorted by LD score. Overall, we estimate that 62% of all
SNPs are associated with a non-zero effect on height. The best-fit line estimates that 3.8% of SNPs have causal effects.
(C) Estimated mean effect size for SNPs, sorted by GIANT p value with the direction (sign) of effect ascertained by GIANT. Replication effect sizes were estimated
using data from the Health and Retirement Study (HRS). The points show averages of 1,000 consecutive SNPS in the p-value-sorted list. The effect size on the
median SNP in the genome is about 10% of that for genome-wide significant hits.
1178 Cell 169, June 15, 2017

true effect size of exactly zero, with SNPs that have a true effect
size that is not zero. Using this approach, we estimated that,
remarkably, 62% of all common SNPs are associated with a
non-zero effect on height (this includes both causal SNPs as
well as nearby SNPs that are correlated through linkage disequi-
librium; Figure 1B). Given that the typical extent of linkage
disequilibrium (LD) is around 10–100 kb (International HapMap
Consortium, 2005), this implies that most 100-kb windows in
the genome include variants that affect height. Stratifying the
ashR analysis by the LD score for each SNP (Bulik-Sullivan
et al., 2015b), we see a clear effect that SNPs with more LD part-
ners are more likely to be associated with height. Under simpli-
fying assumptions (see Supplemental Information), the best-fit
curve suggests that $3.8% of 1000 Genomes SNPs have causal
effects on height.
As validation, we used the regression estimate from each
SNP in the height meta-analysis to predict its direction of ef-
fect on height (Figure 1C) and then examined the extent to
which SNP effects are consistent in a smaller, independent da-
taset from the Health and Retirement Study (Juster and Suz-
man, 1995). In brief, we computed the mean replication effect
sizes of height-increasing alleles as determined by GIANT. Un-
der the null hypothesis of no true signal, the replication effect
sizes would be centered on zero; when there is true signal,
the observed mean effect sizes can be considered a lower
bound on the true effect sizes due to occasional sign errors
in GIANT.
Strikingly, we find clear enrichment of shared directional
signal for most SNPs, even for SNPs with p values as large
as 0.5 (Figure 1C). Across all SNPs genome-wide, the median
SNP is associated with an effect size of 0.14 mm, which is
approximately one-tenth the median effect size of genome-
wide significant SNPs (1.43 mm). We also obtained similar re-
sults starting from a smaller family-based GWAS, confirming
that the signals are not driven by confounding from population
structure (Supplemental Information). Putting the various
lines of evidence together, we estimate that more than
100,000 SNPs exert independent causal effects on height,
similar to an early estimate of 93,000 causal variants based
on a different approach (Goldstein, 2009) (Supplemental Infor-
mation).
In summary, we conclude that there is an extremely large num-
ber of causal variants with tiny effect sizes on height and, more-
over, that these are spread very widely across the genome, such
that most 100-kb windows contribute to variance in height. More
generally, the heritability of complex traits and diseases is
spread broadly across the genome (Loh et al., 2015; Shi et al.,
2016), implying that a substantial fraction of all genes contribute
to variation in disease risk. These observations seem inconsis-
tent with the expectation that complex trait variants are primarily
in specific biologically relevant genes and pathways. To explore
this further, we turn next to data on functional enrichment of
signals.
Enrichment of Genetic Signals in Transcriptionally
Active Regions
As shown above for height, GWAS signals tend to be mark-
edly enriched in predicted gene regulatory elements. In partic-
ular, many groups have shown that disease-associated SNPs
are enriched in active chromatin and particularly in chromatin
that is active in cell types relevant to disease (Trynka et al.,
2013; Farh et al., 2015; Finucane et al., 2015; Kundaje et al.,
2015). Similarly, signals also aggregate near genes that
are expressed in relevant cell types (Hu et al., 2011; Wood
et al., 2014).
An intuitive interpretation is that the cell-type-based regula-
tory maps point us toward cell-type-specific regulatory ele-
ments that control specific functions of those cells and thereby
drive disease biology. Indeed, the relevant papers often
describe these analyses as highlighting ‘‘cell-type-specific’’ as-
pects of regulation. But given that the heritability signal is so
widespread, we wanted to understand whether the signal is
specifically concentrated in chromatin that is active in just the
relevant (or related) cell types, as opposed to chromatin that
is broadly active.
To explore this question, we used active chromatin data
measured in ten broadly defined cell-type groups (e.g., im-
mune, central nervous system (CNS), cardiovascular, etc.).
A region was considered active in a cell-type group if it was de-
tected as active for any cell type in that group. We applied
stratified LD score regression—a method that estimates how
much different classes of SNPs contribute to heritability (Finu-
cane et al., 2015). We focused on three well-powered GWAS
studies that showed clear enrichment within a single cell-type
group in a previous analysis: Crohn’s disease (immune), rheu-
matoid arthritis (RA, immune), and schizophrenia (CNS) (Finu-
cane et al., 2015).
While there are strong cell-type effects, these are largely inde-
pendent of the breadth of chromatin activity. For example, we
observed that SNPs in chromatin that is broadly active across
most cell types make substantial contributions to heritability.
On average, SNPs in broadly active elements contribute roughly
as much to heritability as do SNPs in cell-type-specific active
chromatin (only for RA are these significantly different;
Figure 2A). Meanwhile, SNPs in chromatin that is inactive or is
active only in irrelevant cell types contribute little or no heritabil-
ity, thus providing an important negative control.
For an alternative viewpoint, we also considered breadth of
gene expression. We estimated the contribution of SNPs in or
near exons for genes with different expression profiles. Based
on GTEx data, we identified genes that are particularly highly
expressed in particular tissue groups, as well as broadly ex-
pressed genes (GTEx Consortium, 2015). As shown for schizo-
phrenia (Figure 2B), SNPs near genes that are expressed in
the brain contribute substantially to heritability, while genes
that are specifically expressed in other tissues contribute little
or nothing. Perhaps intuitively, SNPs near genes expressed
specifically in brain contribute more heritability per SNP than
SNPs near genes with broad expression profiles. However,
only a modest fraction of all brain-expressed genes are specif-
ically upregulated in brain. Hence, broadly expressed genes
actually contribute more to the overall heritability than do
brain-specific genes.
In summary, genetic contribution to disease is heavily concen-
trated in regions that are transcribed or marked by active
chromatin in relevant tissues, but there is little enrichment for
Cell 169, June 15, 2017 1179

cell-type-specific regulatory elements versus broadly actively
regions. As expected, there appears to be little or no genetic
contribution from regions that are inactive in these tissues. To
investigate the question of GWAS specificity further, we next
examined evidence for enrichment of associated genes in spe-
cific functional categories.
Weak Enrichment of Genetic Signals by Functional
Categories
We considered the contributions of genes from different func-
tional ontologies. As expected, we found that the genetic
signals for the two autoimmune diseases (Crohn’s and RA)
were most enriched in ontologies corresponding to ‘‘immune
response’’ and ‘‘inflammatory response,’’ whereas schizo-
phrenia heritability was most enriched in nervous-system-
related genes with ontologies such as ‘‘ion channel activity’’
Figure 2. Heritability Tends to Be Enriched
in Regions that Are Transcriptionally Active
in Relevant Tissues
(A) Contributions to heritability (relative to random
SNPs) as a function of chromatin context. There is
enrichment for signal among SNPs that are in
chromatin active in the relevant tissue, regardless
of the overall tissue breadth of activity.
(B) Genes with brain-specific expression show the
strongest enrichment of schizophrenia signal (left),
but broadly expressed genes contribute more to
total heritability due to their greater number (right).
and ‘‘calcium ion transport’’ (Figure 3).
However, these enrichments were rela-
tively modest, and for all three diseases,
we observed a strong linear relationship
between the sizes of the functional cate-
gories and the proportion of heritability
that they contributed. Broad functional
categories contribute more total trait her-
itability than do genes in apparently dis-
ease-relevant functional categories, and
for all three diseases, the largest contrib-
utor to heritability was simply the largest
category, namely protein binding.
Moreover, these results are markedly
different from analysis of rare variants
implicated in schizophrenia. Recent
studies of rare variants have consistently
found enrichment of synaptic genes and
other gene sets involved in neuronal func-
tions within de novo, rare, and CNV poly-
morphism sets (Table 1). In contrast, anal-
ysis of the 108 genome-wide significant
loci from GWAS found examples of hits in
relevant genes but no ontology categories
that were significant overall (Ripke et al.,
2014), consistent with the weak enrich-
ment described above for the heritability
analysis of the same data. Together, these
results suggest that the types of genes de-
tectedin rare variant studies—whichcandetecthighlydeleterious
variants with large effect sizes—play more direct roles in schizo-
phrenia than do genes identified from GWAS based on common
variants.
An Extended Model for Complex Traits
In summary, for a variety of traits, the largest-effect variants are
modestly enriched in specific genes or pathways that may play
direct roles in disease. However, the SNPs that contribute the
bulk of the heritability tend to be spread across the genome
and are not near genes with disease-specific functions. The
clearest pattern is that the association signal is broadly en-
riched in regions that are transcriptionally active or involved in
transcriptional regulation in disease-relevant cell types but ab-
sent from regions that are transcriptionally inactive in those cell
types. For typical traits, huge numbers of variants contribute to
1180 Cell 169, June 15, 2017

heritability, in striking consistency with Fisher’s century-old
infinitesimal model.
To make sense of these observations, we propose an ‘‘omni-
genic’’ model of complex traits (Figure 4). First, we assume that
most traits can be directly affected by a modest number of genes
or gene pathways with specific roles in disease etiology, as well
as their direct regulators (Chakravarti and Turner, 2016). We refer
to these as ‘‘core genes.’’ Such genes will tend to have biologi-
cally interpretable roles in disease, such as the roles of IRX3 and
IRX5 in controlling adipocyte differentiation, with consequent ef-
fects on obesity (Claussnitzer et al., 2015), or the role of the C4
genes on synaptic pruning in development, thereby affecting
schizophrenia risk (Sekar et al., 2016). Furthermore, when core
genes are damaged by loss of function or other particularly
damaging mutations, we can anticipate that these will tend to
have the strongest effects on disease risk (although the actual
degree of increased risk conferred by the largest effect-size mu-
tations varies greatly across traits; Krumm et al., 2015; Marouli
et al., 2017). In practice, the sorting of core genes from peripheral
genes may be on a graduated scale, as opposed to a binary clas-
sification.
Second, we need to understand why core genes generally
contribute just a small part of the total heritability and how
most genes expressed in relevant cell types could make non-
zero contributions to heritability. To resolve this, we propose
that cell regulatory networks are highly interconnected to the
extent that any expressed gene is likely to affect the regulation
or function of core genes.
At this time, our understanding of cellular regulatory networks
remains incomplete, but the relevant connections likely include
all layers of interactions among cellular molecules, including
transcriptional networks, post-translational modifications, pro-
tein-protein interactions, and intercellular signaling (Furlong,
2013). In particular cases, it has been possible to elucidate the
most important wiring connections in gene regulatory networks
that drive development or disease (Davidson, 2010; Chatterjee
et al., 2016). However, we still have very limited knowledge of
how weaker effects such as expression QTLs percolate through
the entire regulatory network. Nonetheless, research in network
theory finds that most real-world networks tend to be highly in-
terconnected; this is referred to as the ‘‘small world’’ property
of networks (Watts and Strogatz, 1998; Strogatz, 2001). Specif-
ically, many kinds of networks have structures consisting of
distinct modules of connected nodes but also frequent long-
range connections. Under these conditions, any two nodes in
the graph are usually connected by just a few steps.
If this is the case in cellular networks, then any gene that is ex-
pressed in a disease-relevant tissue is likely to be just a few steps
from one or more core genes. Consequently, any variant that af-
fects expression of a ‘‘peripheral’’ gene is likely to have non-zero
effects on regulation of the core genes and thereby incur a small
effect on disease risk. Crucially, because the total set of ex-
pressed genes may outnumber core genes by 100:1 or more,
the sum of small effects across peripheral genes can far exceed
the genetic contribution of variants directly affecting the core
genes themselves.
Our model posits that information flows from regulatory var-
iants, e.g., by affecting chromatin activity, to cis regulation of
nearby genes and ultimately to affect the activity of other
genes. cis-eQTLs (cis-acting expression quantitative trait
loci) may in turn affect mRNA or protein levels of other un-
linked genes via the regulatory network (i.e., the variants
would also be trans-acting eQTLs for genes elsewhere in the
genome) but might also affect other functions such as post-
translational modification or subcellular localization. At pre-
sent, detection of trans-QTLs is challenging in current sample
Figure 3. Gene Ontology Enrichments for Three Diseases, with Categories of Particular Interest Labeled
The x axis indicates the fraction of SNPs in each category; the y axis shows the fraction of heritability assigned to each category as a fraction of the heritability
assigned to all SNPs. Note that the diagonal indicates the genome-wide average across all SNPs; most GO categories lie above the line due to the general
enrichment of signal in and around genes. Analysis by stratified LD score regression (Finucane et al., 2015).
Cell 169, June 15, 2017 1181

sizes (Westra et al., 2013; Jo et al., 2016), but it is estimated
that $70% of mRNA heritability is determined by trans-acting
factors (Price et al., 2011). Moreover, many trans-QTLs may
act through protein networks and thus may not be detectable
from RNA, though current data on trans-acting controls of pro-
teins are very limited (Battle et al., 2015; Chick et al., 2016;
Sun et al., 2017).
Lastly, many diseases are mediated through multiple cell
types—for example, different immune cell subsets for autoim-
mune disease or even unrelated tissues such as brain and adi-
pose tissue for obesity. Furthermore, although GWAS hits are
highly enriched in active chromatin, only a modest fraction can
currently be explained by known eQTLs (Chun et al., 2017). This
gap may imply that many risk variants affect expression only in
narrowly defined cell types or under precise conditions such as
immune stimulation (Alasoo et al., 2017). When disease risk is
mediated through multiple cell types or highly specialized cell
types, we anticipate that the cellular networks would vary across
cell types (Price et al., 2011; Sonawane et al., 2017). The quanti-
tative effect of any given variant would then be an average of its
effect size in each cell type, weighted by cell type importance.
In summary, the omnigenic model of complex disease pro-
poses that essentially any gene with regulatory variants in at
least one tissue that contributes to disease pathogenesis is likely
to have nontrivial effects on risk for that disease. Furthermore,
the relative effect sizes are such that, since core genes are
hugely outnumbered by peripheral genes, a large fraction of
the total genetic contribution to disease comes from peripheral
genes that do not play direct roles in disease.
Widespread Pleiotropy
There has recently been considerable interest in identifying
particular variants with pleiotropic effects on different traits (Cot-
sapas et al., 2011; Pickrell et al., 2016) as well as in identifying
pairs of traits with correlated genetic effects (Bulik-Sullivan
et al., 2015a). However, the observation that genetic signals
are spread broadly across the genome implies that pleiotropy
may be ubiquitous (Visscher and Yang, 2016).
Indeed, the omnigenic model predicts that virtually any variant
with regulatory effects in a given tissue is likely to have (weak) ef-
fects on all diseases that are modulated through that tissue.
Many eQTLs are active in all tissues, and consequently these
may have weak effects on most or even all traits.
We refer to this form of pleiotropy as ‘‘network pleiotropy,’’ i.e.,
the principle that a single variant may affect multiple traits
because those traits are mediated through the same cell type(s)
and hence regulated through the same network(s)—and not
because the traits are directly causally related. Traits that share
core genes or whose genes are close in the network will tend to
have correlated effects. Conversely, traits that are mediated
through the same tissue but have no overlap of core genes
may show little or no correlation in effects even though many
causal variants are shared.
If network pleiotropy is widespread, this raises challenges
for the interpretation of genetic correlations and for Mendelian
Randomization studies (Bulik-Sullivan et al., 2015a; Davey Smith
and Hemani, 2014). Mendelian Randomization generally as-
sumes that pleiotropy between traits that are not causally
related—also referred to as ‘‘type I pleiotropy’’ (Wagner and
Zhang, 2011)—is rare. It remains to be determined whether the
effects of network pleiotropy would be strong enough to drive
significant signals in practice, especially if the core genes are
far apart in the network.
Evolutionary Change of Complex Traits
The observation that many traits are affected by huge numbers
of variants also has important implications for studies of evolu-
tionary change. Within the evolutionary community, there has
been great interest in identifying particular genetic variants that
are responsible for adaptive changes, both within and between
species (Vitti et al., 2013). While this work has produced a num-
ber of interesting examples, we argue that these are not likely to
be representative of most evolutionary change. Instead, most
adaptive changes may proceed by polygenic adaptation, i.e.,
species adapt by small allele frequency shifts of many causal
variants across the genome (Pritchard et al., 2010). For example,
if 105
variants affect height by 0.15 mm each, then even a small
shift in average allele frequencies could generate a large shift in
average height; e.g., a 0.5% genome-wide increase in the fre-
quency of ‘‘tall’’ alleles would generate a 15 cm shift in average
height. There is now a growing collection of examples of recent
polygenic adaptation in humans, especially for morphometric
Table 1. Summary of Gene Sets that Show Functional Enrichment in Recent Large-Scale Papers on Schizophrenia
Variant Type Gene Set/Ontology Enrichment P Value Reference
Rare ARC p = 1.6 3 10À3
Purcell et al. (2014)
voltage-gated calcium channel p = 1.9 3 10À3
de novo ARC p = 4.8 3 10À4
Fromer et al. (2014)
N-methyl-D-aspartate receptor (NMDAR) p = 2.5 3 10À2
CNV ARC p = 1.8 3 10À4
The Psychiatric Genetics Consortium (2016)
Synaptic gene p = 2.8 3 10À11
GWAS glutamatergic neurotransmission not significanta
Ripke et al. (2014)
synaptic plasticity
Studies of rare and de novo variants and CNVs—which tend to identify larger-effect variants—show clearer evidence of enrichment than seen in
GWAS. The p values are shown without multiple testing correction, but corrected p values are <0.05.
a
Consistent with studies of rare variants, Ripke et al. (2014) identified associated loci near several genes involved in glutamatergic neurotransmission
and synaptic plasticity, but these categories did not show a statistically significant enrichment for GWAS hits. ARC: activity-regulated cytoskeleton-
associated scaffold protein.
1182 Cell 169, June 15, 2017

traits including height, BMI, and infant birth size (Turchin et al.,
2012; Field et al., 2016).
We anticipate that many of the more dramatic phenotypic dif-
ferences seen between species are also driven by an accumula-
tion of tiny effects and that larger-effect differences are likely to
be exceptions to the rule. For example, there are $40 million sin-
gle-nucleotide differences between humans and chimpanzees. If
1% of these affect chromatin function or other aspects of regu-
lation, then there could easily be a half-million differences be-
tween the two species with small but nonzero effects on pheno-
types (these need not all be adaptive), and these would likely
dominate the contributions of a handful of large-effect loci.
Turning to the within-species level, one important open ques-
tion is whether pleiotropic effects limit how many traits can be
selected for at once. As described above, pleiotropy is likely
ubiquitous in the genome. This may place constraints on the abil-
ity of selection to shift allele frequencies, as a change in the fre-
quency of one variant must be balanced by changes at other
sites. Does this effectively limit the number of independent poly-
genic traits that can be simultaneously selected? There has been
previous consideration of the extent to which pleiotropy shapes
Figure 4. An Omnigenic Model of Complex
Traits
(A) For any given disease phenotype, a limited
number of genes have direct effects on disease
risk. However, by the small world property of
networks, most expressed genes are only a few
steps from the nearest core gene and thus may
have non-zero effects on disease. Since core
genes only constitute a tiny fraction of all genes,
most heritability comes from genes with indirect
effects.
(B) Diseases are generally associated with
dysfunction of speciﬁc tissues; genetic variants
are only relevant if they perturb gene expression
(and hence network state) in those tissues. For
traits that are mediated through multiple cell
types or tissues, the overall effect size of any
given SNP would be a weighted average of its
effects in each cell type.
variation and adaptation (Barton, 1990;
Walsh and Blows, 2009), but we believe
this area is ripe for further exploration in
the light of modern data.
Future Directions
Huge numbers of genes contribute to the
heritability for complex diseases. This
fact raises fundamental questions about
how genetic variation perturbs genetic
systems to produce phenotypes. We
have proposed one possible model, and
it will be important to test this and
perhaps others. There are deep chal-
lenges to fully understanding the impact
of very small effects in organismal sys-
tems, so we believe there is great need
to develop cell-based model systems
that can recapitulate aspects of complex
traits. Furthermore, we still have limited understanding of
cellular networks, and it will be important to develop highly pre-
cise, high-throughput techniques for mapping networks in
diverse cell types, especially at the protein level. We suggest
the following key questions and tests of the omnigenic model:
d For a variety of representative traits: How many distinct
variants and how many genes contribute causal variation?
What fraction of this variation is in non-core genes? Which
traits are closer to (or further from) the omnigenic extreme?
d Are there variants that affect expression in the cell types
that drive a particular disease but have no effect on disease
risk? While traits vary in terms of the importance of the
largest-effect variants, the strongest form of the omnigenic
model predicts that essentially all regulatory variants active
in relevant cell types would contribute non-zero effects.
d If most genetic variants act through cellular networks,
then what mediates these connections? Transcriptional
regulation, post-translational modiﬁcation, protein-
protein interaction, and intercellular signaling may all
contribute. What is the nature and frequency of long-range
Cell 169, June 15, 2017 1183

interactions in cellular networks? How do network archi-
tectures vary across cell types and tissues?
d As we get increasingly precise measurements of the
percolation of genetic variation through cellular networks,
can we infer the effects of peripheral genes from their rela-
tion to core genes?
d Is the conceptual distinction between core genes and pe-
ripheral genes useful for understanding disease, and if so,
how should core genes be defined? One possible formal
definition is that, conditional on the genotype and expres-
sion levels of all core genes, the genotypes and expression
levels of peripheral genes no longer matter. Less formally,
we might think of core genes as the genes that (if mutated
or deleted) have the strongest effects, as seen for large-ef-
fect mutations in autism (Krumm et al., 2015). Or we might
think of core genes simply as the genes with interpretable
mechanistic links to disease. Alternatively, some diseases
may not even have core genes—instead, the global activity
of all genes might help to set cellular system states that
determine cellular function and disease risk (Preininger
et al., 2013).
Our model also raises questions about the next generation of
mapping studies. One goal of gene mapping is to identify core
genes and pathways that drive disease. These provide mecha-
nistic insights into disease biology and may suggest druggable
targets. The biggest hits from GWAS have helped to pinpoint
important core genes. After these have been found, the next
most promising step is to hunt for lower-frequency variants of
larger effects, which likely contribute little to heritability but
may implicate additional core genes. Deep sequencing has
not been uniformly successful for all traits (possibly due to
insufficient sample sizes; Marouli et al., 2017), but following
the identification of the biggest association hits among com-
mon variants, large-scale sequencing is the most promising
next step. In the short-term, exome sequencing is likely the
most cost-effective approach, given current evidence that
larger-effect variants are more likely to affect protein-coding
sequences.
Nonetheless, large-scale genotyping data will continue to be
valuable for two reasons. First, very deep association data will
be essential for developing personalized risk prediction. Second,
these data will be essential for modeling the flow of regulatory in-
formation through cellular networks. For a complete understand-
ing of disease genetics, we will want to know why increased
expression of gene X increases risk for diseases Y and Z. For
this, we will need to understand cellular networks much better
and to have estimates of disease risk in very large samples.
In summary, many complex traits are driven by enormously
large numbers of variants of small effects, potentially implicating
most regulatory variants that are active in disease-relevant tis-
sues. To explain these observations, we propose that disease
risk is largely driven by genes with no direct relevance to disease
and is propagated through regulatory networks to a much
smaller number of core genes with direct effects. If this model
is correct, then it implies that detailed mapping of cell-specific
regulatory networks will be an essential task for fully understand-
ing human disease biology.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Materials and Methods, five figures, and
one table and can be found with this article online at http://dx.doi.org/10.
1016/j.cell.2017.05.038.
ACKNOWLEDGMENTS
This work was supported by RO1 HG008140, the National Science Foundation
graduate research fellowship program, and the Howard Hughes Medical Insti-
tute. We thank many colleagues for helpful conversations or comments,
including D. Golan, W. Greenleaf, A. Harpak, A. Marson, J. Pickrell, M. Prze-
worski, G. Sella, and three anonymous reviewers.
REFERENCES
Alasoo, K., Rodrigues, J., Mukhopadhyay, S., Knights, A.J., Mann, A.L.,
Kundu, K., HIPSCI Consortium, Hale, C., Dougan, G., and Gaffney, D.J.
(2017). Genetic effects on chromatin accessibility foreshadow gene expres-
sion changes in macrophage immune response. bioRxiv, https://doi.org/10.
1101/102392.
Barton, N.H. (1990). Pleiotropic models of quantitative variation. Genetics 124,
773–782.
Barton, N.H., Etheridge, A.M., and Veber, A. (2016). The infinitesimal model.
bioRxiv, https://doi.org/10.1101/039768.
Battle, A., Khan, Z., Wang, S.H., Mitrano, A., Ford, M.J., Pritchard, J.K., and
Gilad, Y. (2015). Genomic variation. Impact of regulatory variation from RNA
to protein. Science 347, 664–667.
Botstein, D., and Risch, N. (2003). Discovering genotypes underlying human
phenotypes: past successes for mendelian disease, future approaches for
complex disease. Nat. Genet. 33 (Suppl), 228–237.
Bulik-Sullivan, B., Finucane, H.K., Anttila, V., Gusev, A., Day, F.R., Loh, P.-R.,
Duncan, L., Perry, J.R., Patterson, N., Robinson, E.B., et al.; ReproGen Con-
sortium; Psychiatric Genomics Consortium; Genetic Consortium for Anorexia
Nervosa of the Wellcome Trust Case Control Consortium 3 (2015a). An atlas of
genetic correlations across human diseases and traits. Nat. Genet. 47,
1236–1241.
Bulik-Sullivan, B.K., Loh, P.R., Finucane, H.K., Ripke, S., Yang, J., Patterson,
N., Daly, M.J., Price, A.L., and Neale, B.M.; Schizophrenia Working Group of
the Psychiatric Genomics Consortium (2015b). LD Score regression distin-
guishes confounding from polygenicity in genome-wide association studies.
Nat. Genet. 47, 291–295.
Califano, A., Butte, A.J., Friend, S., Ideker, T., and Schadt, E. (2012).
Leveraging models of cell regulation and GWAS data in integrative network-
based association studies. Nat. Genet. 44, 841–847.
Chakravarti, A., and Turner, T.N. (2016). Revealing rate-limiting steps in com-
plex disease biology: The crucial importance of studying rare, extreme-pheno-
type families. BioEssays 38, 578–586.
Chatterjee, S., Kapoor, A., Akiyama, J.A., Auer, D.R., Lee, D., Gabriel, S., Ber-
rios, C., Pennacchio, L.A., and Chakravarti, A. (2016). Enhancer Variants Syn-
ergistically Drive Dysfunction of a Gene Regulatory Network In Hirschsprung
Disease. Cell 167, 355–368.e10.
Chick, J.M., Munger, S.C., Simecek, P., Huttlin, E.L., Choi, K., Gatti, D.M., Ra-
ghupathy, N., Svenson, K.L., Churchill, G.A., and Gygi, S.P. (2016). Defining
the consequences of genetic variation on a proteome-wide scale. Nature
534, 500–505.
Chun, S., Casparino, A., Patsopoulos, N.A., Croteau-Chonka, D.C., Raby,
B.A., De Jager, P.L., Sunyaev, S.R., and Cotsapas, C. (2017). Limited statisti-
cal evidence for shared genetic effects of eQTLs and autoimmune-disease-
associated loci in three major immune-cell types. Nat. Genet. 49, 600–605.
Claussnitzer, M., Dankel, S.N., Kim, K.H., Quon, G., Meuleman, W., Haugen,
C., Glunk, V., Sousa, I.S., Beaudry, J.L., Puviindran, V., et al. (2015). FTO
Obesity Variant Circuitry and Adipocyte Browning in Humans. N. Engl. J.
Med. 373, 895–907.
1184 Cell 169, June 15, 2017

Cotsapas, C., Voight, B.F., Rossin, E., Lage, K., Neale, B.M., Wallace, C., Abe-
casis, G.R., Barrett, J.C., Behrens, T., Cho, J., et al.; FOCiS Network of Con-
sortia (2011). Pervasive sharing of genetic effects in autoimmune disease.
PLoS Genet. 7, e1002254.
Davey Smith, G., and Hemani, G. (2014). Mendelian randomization: genetic
anchors for causal inference in epidemiological studies. Hum. Mol. Genet.
23 (R1), R89–R98.
Davidson, E.H. (2010). Emerging properties of animal gene regulatory net-
works. Nature 468, 911–920.
De Rubeis, S., He, X., Goldberg, A.P., Poultney, C.S., Samocha, K., Cicek,
A.E., Kou, Y., Liu, L., Fromer, M., Walker, S., et al.; DDD Study; Homozygosity
Mapping Collaborative for Autism; UK10K Consortium (2014). Synaptic, tran-
scriptional and chromatin genes disrupted in autism. Nature 515, 209–215.
Farh, K.K.-H., Marson, A., Zhu, J., Kleinewietfeld, M., Housley, W.J., Beik, S.,
Shoresh, N., Whitton, H., Ryan, R.J., Shishkin, A.A., et al. (2015). Genetic and
epigenetic fine mapping of causal autoimmune disease variants. Nature 518,
337–343.
Field, Y., Boyle, E.A., Telis, N., Gao, Z., Gaulton, K.J., Golan, D., Yengo, L., Ro-
cheleau, G., Froguel, P., McCarthy, M.I., and Pritchard, J.K. (2016). Detection
of human adaptation during the past 2000 years. Science 354, 760–764.
Finucane, H.K., Bulik-Sullivan, B., Gusev, A., Trynka, G., Reshef, Y., Loh, P.-R.,
Anttila, V., Xu, H., Zang, C., Farh, K., et al.; ReproGen Consortium; Schizo-
phrenia Working Group of the Psychiatric Genomics Consortium; RACI Con-
sortium (2015). Partitioning heritability by functional annotation using
genome-wide association summary statistics. Nat. Genet. 47, 1228–1235.
Fisher, R.A. (1918). The correlation between relatives on the supposition of
Mendelian inheritance. Trans. R. Soc. Edinb. 52, 399–433.
Fromer, M., Pocklington, A.J., Kavanagh, D.H., Williams, H.J., Dwyer, S.,
Gormley, P., Georgieva, L., Rees, E., Palta, P., Ruderfer, D.M., et al. (2014).
De novo mutations in schizophrenia implicate synaptic networks. Nature
506, 179–184.
Furlong, L.I. (2013). Human diseases through the lens of network biology.
Trends Genet. 29, 150–159.
Goldstein, D.B. (2009). Common genetic variation and human traits. N. Engl.
J. Med. 360, 1696–1698.
GTEx Consortium (2015). Human genomics. The Genotype-Tissue Expression
(GTEx) pilot analysis: multitissue gene regulation in humans. Science 348,
648–660.
Hu, X., Kim, H., Stahl, E., Plenge, R., Daly, M., and Raychaudhuri, S. (2011).
Integrating autoimmune risk loci with gene-expression data identifies specific
pathogenic immune cell subsets. Am. J. Hum. Genet. 89, 496–506.
International HapMap Consortium (2005). A haplotype map of the human
genome. Nature 437, 1299–1320.
Jo, B., He, Y., Strober, B.J., Parsana, P., Aguet, F., Brown, A.A., Castel, S.E.,
Gamazon, E.R., Gewirtz, A., Gliner, G., et al. (2016). Distant regulatory effects
of genetic variation in multiple human tissues. bioRxiv, https://doi.org/10.
1101/074419.
Jostins, L., Ripke, S., Weersma, R.K., Duerr, R.H., McGovern, D.P., Hui, K.Y.,
Lee, J.C., Schumm, L.P., Sharma, Y., Anderson, C.A., et al.; International IBD
Genetics Consortium (IIBDGC) (2012). Host-microbe interactions have shaped
the genetic architecture of inflammatory bowel disease. Nature 491, 119–124.
Juster, F.T., and Suzman, R. (1995). An overview of the Health and Retirement
Study. J. Hum. Resour. 30, S7–S56.
Krumm, N., Turner, T.N., Baker, C., Vives, L., Mohajeri, K., Witherspoon, K.,
Raja, A., Coe, B.P., Stessman, H.A., He, Z.-X., et al. (2015). Excess of rare, in-
herited truncating mutations in autism. Nat. Genet. 47, 582–588.
Kundaje, A., Meuleman, W., Ernst, J., Bilenky, M., Yen, A., Heravi-Moussavi,
A., Kheradpour, P., Zhang, Z., Wang, J., Ziller, M.J., et al.; Roadmap
Epigenomics Consortium (2015). Integrative analysis of 111 reference human
epigenomes. Nature 518, 317–330.
Li, Y.I., van de Geijn, B., Raj, A., Knowles, D.A., Petti, A.A., Golan, D., Gilad, Y.,
and Pritchard, J.K. (2016). RNA splicing is a primary link between genetic vari-
ation and disease. Science 352, 600–604.
Locke, A.E., Kahali, B., Berndt, S.I., Justice, A.E., Pers, T.H., Day, F.R., Powell,
C., Vedantam, S., Buchkovich, M.L., Yang, J., et al.; LifeLines Cohort Study;
ADIPOGen Consortium; AGEN-BMI Working Group; CARDIOGRAMplusC4D
Consortium; CKDGen Consortium; GLGC; ICBP; MAGIC Investigators;
MuTHER Consortium; MIGen Consortium; PAGE Consortium; ReproGen Con-
sortium; GENIE Consortium; International Endogene Consortium (2015). Ge-
netic studies of body mass index yield new insights for obesity biology. Nature
518, 197–206.
Loh, P.-R., Bhatia, G., Gusev, A., Finucane, H.K., Bulik-Sullivan, B.K., Pollack,
S.J., de Candia, T.R., Lee, S.H., Wray, N.R., Kendler, K.S., et al.; Schizophrenia
Working Group of Psychiatric Genomics Consortium (2015). Contrasting ge-
netic architectures of schizophrenia and other complex diseases using fast
variance-components analysis. Nat. Genet. 47, 1385–1392.
Manolio, T.A., Collins, F.S., Cox, N.J., Goldstein, D.B., Hindorff, L.A., Hunter,
D.J., McCarthy, M.I., Ramos, E.M., Cardon, L.R., Chakravarti, A., et al.
(2009). Finding the missing heritability of complex diseases. Nature 461,
747–753.
Marouli, E., Graff, M., Medina-Gomez, C., Lo, K.S., Wood, A.R., Kjaer, T.R.,
Fine, R.S., Lu, Y., Schurmann, C., Highland, H.M., et al.; EPIC-InterAct Con-
sortium; CHD Exome+ Consortium; ExomeBP Consortium; T2D-Genes Con-
sortium; GoT2D Genes Consortium; Global Lipids Genetics Consortium;
ReproGen Consortium; MAGIC Investigators (2017). Rare and low-frequency
coding variants alter human adult height. Nature 542, 186–190.
Maurano, M.T., Humbert, R., Rynes, E., Thurman, R.E., Haugen, E., Wang, H.,
Reynolds, A.P., Sandstrom, R., Qu, H., Brody, J., et al. (2012). Systematic
localization of common disease-associated variation in regulatory DNA. Sci-
ence 337, 1190–1195.
Pickrell, J.K. (2014). Joint analysis of functional genomic data and genome-
wide association studies of 18 human traits. Am. J. Hum. Genet. 94, 559–573.
Pickrell, J.K., Berisa, T., Liu, J.Z., Se´ gurel, L., Tung, J.Y., and Hinds, D.A.
(2016). Detection and interpretation of shared genetic influences on 42 human
traits. Nat. Genet. 48, 709–717.
Preininger, M., Arafat, D., Kim, J., Nath, A.P., Idaghdour, Y., Brigham, K.L., and
Gibson, G. (2013). Blood-informative transcripts define nine common axes of
peripheral blood gene expression. PLoS Genet. 9, e1003362.
Price, A.L., Helgason, A., Thorleifsson, G., McCarroll, S.A., Kong, A., and Ste-
fansson, K. (2011). Single-tissue and cross-tissue heritability of gene expres-
sion via identity-by-descent in related or unrelated individuals. PLoS Genet.
7, e1001317.
Pritchard, J.K., Pickrell, J.K., and Coop, G. (2010). The genetics of human
adaptation: hard sweeps, soft sweeps, and polygenic adaptation. Curr. Biol.
20, R208–R215.
Purcell, S.M., Wray, N.R., Stone, J.L., Visscher, P.M., O’Donovan, M.C., Sulli-
van, P.F., Sklar, P., Ruderfer, D.M., McQuillin, A., Morris, D.W., et al.; Interna-
tional Schizophrenia Consortium (2009). Common polygenic variation contrib-
utes to risk of schizophrenia and bipolar disorder. Nature 460, 748–752.
Purcell, S.M., Moran, J.L., Fromer, M., Ruderfer, D., Solovieff, N., Roussos, P.,
O’Dushlaine, C., Chambert, K., Bergen, S.E., Ka¨ hler, A., et al. (2014). A poly-
genic burden of rare disruptive mutations in schizophrenia. Nature 506,
185–190.
Ripke, S., Neale, B.M., Corvin, A., Walters, J.T., Farh, K.-H., Holmans, P.A.,
Lee, P., Bulik-Sullivan, B., Collier, D.A., Huang, H., et al.; Schizophrenia Work-
ing Group of the Psychiatric Genomics Consortium (2014). Biological insights
from 108 schizophrenia-associated genetic loci. Nature 511, 421–427.
Risch, N., Spiker, D., Lotspeich, L., Nouri, N., Hinds, D., Hallmayer, J., Kalayd-
jieva, L., McCague, P., Dimiceli, S., Pitts, T., et al. (1999). A genomic screen of
autism: evidence for a multilocus etiology. Am. J. Hum. Genet. 65, 493–507.
Sekar, A., Bialas, A.R., de Rivera, H., Davis, A., Hammond, T.R., Kamitaki, N.,
Tooley, K., Presumey, J., Baum, M., Van Doren, V., et al.; Schizophrenia Work-
ing Group of the Psychiatric Genomics Consortium (2016). Schizophrenia risk
from complex variation of complement component 4. Nature 530, 177–183.
Cell 169, June 15, 2017 1185

Shi, H., Kichaev, G., and Pasaniuc, B. (2016). Contrasting the genetic architec-
ture of 30 complex traits from summary association data. Am. J. Hum. Genet.
99, 139–153.
Simons, Y.B., Turchin, M.C., Pritchard, J.K., and Sella, G. (2014). The delete-
rious mutation load is insensitive to recent population history. Nat. Genet. 46,
220–224.
Smemo, S., Tena, J.J., Kim, K.-H., Gamazon, E.R., Sakabe, N.J., Go´ mez-
Marıń, C., Aneas, I., Credidio, F.L., Sobreira, D.R., Wasserman, N.F., et al.
(2014). Obesity-associated variants within FTO form long-range functional
connections with IRX3. Nature 507, 371–375.
Sonawane, A.R., Platig, J., Fagny, M., Chen, C.-Y., Paulson, J.N., Lopes-Ra-
mos, C.M., DeMeo, D.L., Quackenbush, J., Glass, K., and Kuijjer, M.L. (2017).
Understanding tissue-specific gene regulation. bioRxiv, https://doi.org/10.
1101/110601.
Stephens, M. (2017). False discovery rates: a new deal. Biostatistics 18,
275–294.
Strogatz, S.H. (2001). Exploring complex networks. Nature 410, 268–276.
Sullivan, P.F., Agrawal, A., Bulik, C., Andreassen, O.A., Borglum, A., Breen, G.,
Cichon, S., Edenberg, H., Faraone, S.V., Gelernter, J., Mathews, C.A., Niever-
gelt, C.M., Smoller, J., and O’Donovan, M. (2017). Psychiatric Genomics: An
Update and an Agenda. bioRxiv, https://doi.org/10.1101/115600.
Sun, B.B., Maranville, J.C., Peters, J.E., Stacey, D., Staley, J.R., Blackshaw, J.,
Burgess, S., Jiang, T., Paige, E., Surendran, P., et al. (2017). Consequences Of
Natural Perturbations In The Human Plasma Proteome. bioRxiv. https://doi.
org/10.1101/134551.
The Psychiatric Genetics Consortium (2016). Contribution of copy number var-
iants to schizophrenia from a genome-wide study of 41,321 subjects. Nat.
Genet. 49, 27–35.
Trynka, G., Sandor, C., Han, B., Xu, H., Stranger, B.E., Liu, X.S., and Ray-
chaudhuri, S. (2013). Chromatin marks identify critical cell types for fine map-
ping complex trait variants. Nat. Genet. 45, 124–130.
Turchin, M.C., Chiang, C.W., Palmer, C.D., Sankararaman, S., Reich, D., and
Genetic Investigation of Anthropometric Traits Consortium, and Hirschhorn,
J.N. (2012). Evidence of widespread selection on standing variation in Europe
at height-associated SNPs. Nat. Genet 44, 1015–1019.
Visscher, P.M., and Yang, J. (2016). A plethora of pleiotropy across complex
traits. Nat. Genet. 48, 707–708.
Visscher, P.M., Medland, S.E., Ferreira, M.A., Morley, K.I., Zhu, G., Cornes,
B.K., Montgomery, G.W., and Martin, N.G. (2006). Assumption-free estimation
of heritability from genome-wide identity-by-descent sharing between full sib-
lings. PLoS Genet. 2, e41.
Vitti, J.J., Grossman, S.R., and Sabeti, P.C. (2013). Detecting natural selection
in genomic data. Annu. Rev. Genet. 47, 97–120.
Wagner, G.P., and Zhang, J. (2011). The pleiotropic structure of the genotype-
phenotype map: the evolvability of complex organisms. Nat. Rev. Genet. 12,
204–213.
Walsh, B., and Blows, M.W. (2009). Abundant genetic variation + strong selec-
tion = multivariate genetic constraints: A geometric view of adaptation. Annu.
Rev. Ecol. Evol. Syst. 40, 41–59.
Watts, D.J., and Strogatz, S.H. (1998). Collective dynamics of ‘small-world’
networks. Nature 393, 440–442.
Weiner, D.J., Wigdor, E.M., Ripke, S., Walters, R.K., Kosmicki, J.A., Grove, J.,
Samocha, K.E., Goldstein, J., Okbay, A., Bybjerg-Gauholm, J., et al. (2016).
Polygenic transmission disequilibrium confirms that common and rare varia-
tion act additively to create risk for autism spectrum disorders. bioRxiv,
https://doi.org/10.1101/089342.
Welter, D., MacArthur, J., Morales, J., Burdett, T., Hall, P., Junkins, H., Klemm,
A., Flicek, P., Manolio, T., Hindorff, L., and Parkinson, H. (2014). The NHGRI
GWAS Catalog, a curated resource of SNP-trait associations. Nucleic Acids
Res. 42 (Database issue, D1), D1001–D1006.
Westra, H.-J., Peters, M.J., Esko, T., Yaghootkar, H., Schurmann, C., Kettu-
nen, J., Christiansen, M.W., Fairfax, B.P., Schramm, K., Powell, J.E., et al.
(2013). Systematic identification of trans eQTLs as putative drivers of known
disease associations. Nat. Genet. 45, 1238–1243.
Wood, A.R., Esko, T., Yang, J., Vedantam, S., Pers, T.H., Gustafsson, S., Chu,
A.Y., Estrada, K., Luan, J., Kutalik, Z., et al.; Electronic Medical Records and
Genomics (eMEMERGEGE) Consortium; MIGen Consortium; PAGEGE Con-
sortium; LifeLines Cohort Study (2014). Defining the role of common variation
in the genomic and biological architecture of adult human height. Nat. Genet.
46, 1173–1186.
Yang, J., Benyamin, B., McEvoy, B.P., Gordon, S., Henders, A.K., Nyholt, D.R.,
Madden, P.A., Heath, A.C., Martin, N.G., Montgomery, G.W., et al. (2010).
Common SNPs explain a large proportion of the heritability for human height.
Nat. Genet. 42, 565–569.
1186 Cell 169, June 15, 2017

An expanded view of complex traits from polygenic to omnigenic

More Related Content

What's hot (9)

Similar to An expanded view of complex traits from polygenic to omnigenic (20)

More from BARRY STANLEY 2 fasd (20)

Recently uploaded (20)

An expanded view of complex traits from polygenic to omnigenic