The 12,111 SNPs affecting human height come from the largest genome-wide association study of 5.4 million individuals to date

The reason why "rich and handsome" is desirable is probably because these three words simply and crudely summarize the instinctive pursuit
of physical advantages, survival resources and aesthetics.
"Tall" is relatively simple—adult height is easy to measure and undisturbed by personal judgment factors, largely related to genetics, and relatively less influenced by other factors—making height a model trait
that geneticists can use to assess "what role common genetic variants play in determining the genetic and biological structure of human multigenotypes.
" But even geneticists who began studying height 20 years ago aren't sure they'll be able to find the common genetic factors
that influence this trait.

Joel Hirschhorn, a pediatric endocrinology researcher at Boston Children's Hospital, has seen many children identified by their pediatricians as unusually short in stature
.
He often tells parents that their children grow slowly because their children have inherited genes, and although scientists estimate that 80 percent of height changes are caused by genes, they don't know what those genes are
.
"After the 20th patient, I thought, 'Hey, where I work can solve this,'" Hirschhorn, a postdoctoral researcher at the Broad Institute, decided to study height at the time, even though everyone told him that height was a polygenic phenotype — there were too many genes involved to find them
.
Undeterred, Hirschhorn turned to genome-wide association studies (GWAS) — scientists scanning the entire genome of a population to identify connections between
genetic variants and traits.
Over the next 20 years, the GIANT consortium reported as many as 3,290 height-associated SNP variants at 712 sites in a GWAS database of nearly 700,000 sample sizes, but because the study focused on participants of European ancestry, the lack of diversity was a common problem
in genetic research.

Researchers from the Broad Institute and the GIANT Consortium, together with 23andMe, collected data from genome-wide association studies of nearly 5.
4 million individuals from different ancestry – more than six times more than in previous studies, including > 4.
08 million participants of European ancestry (EUR, 75.
8% of the total sample); >470,000 participants of East Asian descent (EAS, 8.
8%); > 450,000 Hispanic participants with typical mixed ancestry (HIS, 8.
5%); >290,000 African Americans of African and European ancestry (AFR, 5.
5%); Nearly 80,000 participants of South Asian descent (SAS, 1.
4%)
.
Using SNP variants cataloged in the HapMap 3 project (HM3), they mapped common genetic associations with adult height and assessed the atlas' saturation with variants, genomic regions, and where possible
.
The analysis confirmed 12,111 common height-related single nucleotide polymorphisms (SNPs), explaining almost all common SNP-based height-related genetic factors
.
These SNPs are clustered in 7,209 nonoverlapping locus fragments with an average size of about 90 kb, covering about 21% of the genome
.
Independently associated SNP densities vary throughout the genome, with regions of high density rich in biologically relevant genes – that is: these 12,111 GWS SNPs are not randomly clustered near each other and near
known bone growth genes.
These 12,111 SNPs (or all SNPs in HapMap 3 panel 2) account for 40% (45%) of the height phenotypic variation in people of European ancestry, but only about
10-20% (14-24%) of other ancestry.
This is the largest study of its kind to date, and overall, it provides a comprehensive map
of specific genomic regions containing the vast majority of common height-related SNP variants.

This atlas is saturated for populations of European ancestry, but further research is needed to achieve equal saturation in other ancestry
.
While the genomic regions known so far to influence height in different populations are the same, including more individual samples of non-European ancestry will be crucial to improve the accuracy of predictions and help identify genetic variants
in specific populations.
"This study covers a wide range of human ancestors and is the largest GWAS to date, but we once again found that further involvement of different ancestors is necessary," said
Yukinori Okada, one of the corresponding authors, a professor at Osaka University School of Medicine and the University of Tokyo, and team leader at Riken Center for Integrative Medicine in Japan.

These SNPs can help researchers develop height prediction tools
for clinical use.
Pediatricians currently predict a child's height based on their family history, but these estimates aren't perfect — for example, they can't predict that a pair of siblings will be different
in height.
SNP-based predictions may be more accurate
.
If doctors notice that your child's height doesn't match predictions, it could be a clue to whether your child has rare hidden diseases that affect growth, such as celiac disease and hormone deficiencies
.

By studying the DNA of more than 5 million people, these teams of geneticists have accomplished what
they thought was impossible years ago.
"We feel this is really a milestone," said Joel Hirschhorn, senior author of the study and a member of the Broad Institute, who is also a professor of genetics and pediatrics
at Boston Children's Hospital and Harvard Medical School.
"We've basically done the work of mapping this genetic influencer to specific genomic regions now, highlighting how increasing sample size can tell us which traits are controlled
by multiple genes.
" The findings, published in the journal Nature, could one day help doctors identify people
who don't reach the height predicted by their genes and may have hidden diseases or defects that affect their growth and health.
The findings also illustrate the power of
genome-wide association studies (GWAS) in revealing the biological basis of disease and, in larger studies, in revealing disease heritability.

"In 2010, we predicted that we should be able to explain 40 per cent of individual height differences, but we didn't expect that it would take 5 million people involving 12,000 DNA mutations and that it would be achieved so quickly," said
Peter Visscher, senior co-author of the study and professor and chair of quantitative genetics at the University of Queensland.
"When GWAS is combined with very large sample sizes, it's a surprisingly powerful experimental design
.
"

As the largest GWAS, the study also provides insights
into how scientists learn from the genome.
Adding participants to GWAS makes it a more reliable and powerful tool, but scientists don't know if there is a tipping point to "saturate" research — that is, when additional data doesn't provide any new insights
.
THE GIANT RESEARCHERS FOUND THAT THE SRPS THEY DETERMINED ULTIMATELY EXPLAINED MORE THAN 90 PERCENT OF THE SNP-BASED VARIATION, SUGGESTING A SATURATION POINT
.
They also found that identifying the approximate contours of biological pathways related to height required less
sample volume than finding precise genomic regions.

Loic Yengo, director of the Statistical Genomics Laboratory at the Institute of Molecular Biosciences at the University of Queensland in Australia, and first and corresponding author of the study, said: "We have been able to use empirical data, rather than previously used theoretical models, to solve long-standing problems
in GWAS research.
"

Geneticists also wondered whether the SNPs it revealed would be distributed across
more and more genomes as GWAS increased.
GIANT'S FINDINGS SHOWED THAT SNPs that affect height are clustered in more than 20 percent of the genome, particularly near genes previously associated with bone growth disorders — for example, 25 SNPs cluster near the ACAN gene, which is mutated
in patients with short stature and skeletal dysplasia.
Some SNPs also hint at signaling pathways that affect bone growth plates, cartilage near the ends of long bones that expand and harden into solid bone
as children grow.

The researchers believe that this clustering of genetic variants may be applicable to other traits and may inform the study of other common diseases, such as high blood pressure or asthma,
which are affected by multiple genes.
NOW THAT THEY KNOW WHICH GENOMIC REGIONS AFFECT HEIGHT, THE GIANT TEAM CAN ALSO BEGIN USING A FINE-MAPPED APPROACH TO TRACK HOW INDIVIDUAL VARIATION AFFECTS HEIGHT
.
Rarer and more complex variations, which may explain heritability that SNPs cannot explain, will also be the target of
future research.

The researchers also discussed the limitations
of the study.
For example, the study focused on SNPs from the HM3 panel, which represent only some of the common genetic variants
.
SNPs outside of the HM3 SNP panel may also be associated
with height.
However, additional information (also known as "hidden heritability") is still likely to be concentrated within the HM3-SNP-based GWS locus, a result that highlights the pervasive allele heterogeneity of height-related locus
.
Another limitation of the study was that more than 75% of the individuals were of European ancestry, failing to address how to identify conditionally independent associations in multi-ancestry studies; Efficient replication analysis
of population-specific genetic associations is not possible.
As with all GWASs, the definitive identification of effector genes and the mechanisms by which genes and variants influence phenotypes remain a key bottleneck
.
Perhaps, breakthrough advances in identifying height causal genes from GWAS can be achieved by combining with larger and larger whole-exome sequencing studies, enabling direct SNP-to-gene mapping
.