GRR: graphical representation of relationship errors.
ABSTRACT A graphical tool for verifying assumed relationships between individuals in genetic studies is described. GRR can detect many common errors using genotypes from many markers. AVAILABILITY: GRR is available at http://bioinformatics.well.ox.ac.uk/GRR.
- [Show abstract] [Hide abstract]
ABSTRACT: Noise-induced hearing loss (NIHL) is a complex disease resulting from the interaction between external and intrinsic/genetic factors. Based on mice studies, one of the most interesting candidate gene for NIHL susceptibility is CDH23-encoding cadherin 23, a component of the stereocilia tip links. The aim of this study was to analyze selected CDH23 single nucleotide polymorphisms (SNPs) and to evaluate their interaction with environmental and individual factors in respect to susceptibility for NIHL in humans. A study group consisted of 314 worst-hearing and 313 best-hearing subjects exposed to occupational noise, selected out of 3,860 workers database. Five SNPs in CDH23 were genotyped using real-time PCR. Subsequently, the main effect of genotype and its interaction with selected environmental and individual factors were evaluated. The significant results within the main effect of genotype were obtained for the SNP rs3752752, localized in exon 21. The effect was observed in particular in the subgroup of young subjects and in those exposed to impulse noise; CC genotype was more frequent among susceptible subjects, whereas genotype CT appeared more often among resistant to noise subjects. The effect of this polymorphism was not modified by none of environmental/individual factors except for blood pressure; however, the latter one should be further investigated. Smoking was shown as an independent factor determining NIHL development. The results of this study confirm that CDH23 genetic variant may modify the susceptibility to NIHL development in humans, as it was earlier proven in mice. Because the differences between the 2 study groups were not necessarily related to susceptibility to noise but they also were prone to age-related cochlear changes, these results should be interpreted with caution until replication in another population.Otology & neurotology: official publication of the American Otological Society, American Neurotology Society [and] European Academy of Otology and Neurotology 02/2014; 35(2):358-65. · 1.44 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Chronic kidney disease (CKD) can be a consequence of diabetes, hypertension, immunologic disorders, and other exposures, as well as genetic factors that are still largely unknown. Glomerular filtration rate (GFR), which is widely used to measure kidney function, has a heritability ranging from 25% to 75%, but only 1.5% of this heritability is explained by genetic loci that have been identified to date. In this study we tested for associations between GFR and 234 SNPs in 26 genes from pathways of blood pressure regulation in 3,025 rural Chinese participants of the "Genetic Epidemiology Network of Salt Sensitivity" (GenSalt) study. We estimated GFR (eGFR) using baseline serum creatinine measurements obtained prior to dietary intervention. We identified significant associations between eGFR and 12 SNPs in 6 genes (ACE, ADD1, AGT, GRK4, HSD11B1, and SCNN1G). The cumulative effect of the protective alleles was an increase in mean eGFR of 4 mL/min per 1.73 m2, while the cumulative effect of the risk alleles was a decrease in mean eGFR of 3 mL/min per 1.73 m2. In addition, we identified a significant interaction between SNPs in CYP11B1 and ADRB2. We have identified common variants in genes from pathways that regulate blood pressure and influence kidney function as measured by eGFR, providing new insights into the genetic determinants of kidney function. Complex genetic effects on kidney function likely involve interactions among genes as we observed for CYP11B1 and ADRB2.PLoS ONE 01/2014; 9(3):e92468. · 3.73 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Non-hereditary colorectal cancer (CRC) is a complex disorder resulting from the combination of genetic and non-genetic factors. Genome-wide association studies (GWAS) are useful for identifying such genetic susceptibility factors. However, the single loci so far associated with CRC only represent a fraction of the genetic risk for CRC development in the general population. Therefore, many other genetic risk variants alone and in combination must still remain to be discovered. The aim of this work was to search for genetic risk factors for CRC, by performing single-locus and two-locus GWAS in the Spanish population.PLoS ONE 01/2014; 9(6):e101178. · 3.73 Impact Factor
BIOINFORMATICS APPLICATIONS NOTE
Vol. 17 no. 8 2001
GRR: graphical representation of relationship
Gonc ¸alo R. Abecasis, Stacey S. Cherny, W. O. C. Cookson and
Lon R. Cardon
Wellcome Trust Centre for Human Genetics, University of Oxford, Roosevelt Drive,
Oxford OX3 7RZ, UK
Received on February 1, 2001; revised on March 27, 2001; accepted on March 28, 2001
Summary: A graphical tool for verifying assumed relation-
ships between individuals in genetic studies is described.
GRR can detect many common errors using genotypes
from many markers.
Availability: GRR is available at http://bioinformatics.well.
Contact: email@example.com; firstname.lastname@example.org
Many large scale linkage and association studies have
been conducted and their popularity is increasing. Simple,
efficient, quality control procedures are essential to
the successful completion of these studies. A common
problem in genetic studies is the misspecification of
relationships between DNA samples (Ott, 1991). Mis-
specification of relationships can lead to inaccurate or
biased results and it is therefore important to verify all
The effects of relationship misspecification are varied.
In studies using family data, problems such as non-
paternity and the mislabeling of monozygotic (MZ) twins
as non-twin full sibs, as well as sample mix-ups can lead
to mistaken inferences about allele sharing. For example,
MZ twins will always share more alleles than other
sibling pairs while, on average, half-siblings will share
fewer alleles than full-siblings. In larger pedigrees, the
potential for relationship misspecification is greater and
the detection of these problems is even harder.
In studies using samples of unrelated individuals, such
as association and pharmacogenetic applications, the pres-
ence of related individuals can lead to a misleading infer-
ence about statistical significance. For example, although
to a certain drug share a certain genotype, the finding is
less striking if some of the individuals are related.
The correct genetic relationship between any two
individuals defines an expected pattern of allele sharing
between them. The details of this pattern can be com-
plex, and will depend on the exact type of relationship,
marker characteristics, population history and inbreeding.
Statistics for verifying relationships through patterns of
allele sharing have been proposed, with various degrees
of sophistication, computing time requirements and
assumptions (Boehnke and Cox, 1997; Goring and Ott,
1997; Broman and Weber, 1998; Epstein et al., 2000;
McPeek and Sun, 2000). Here we describe a simple,
general approach for verifying that individuals with the
same specified relationship have similar patterns of allele
sharing. Unlike other approaches, our method does not
require specification of allele frequencies or any other
population parameter. It is expected to be robust to a small
level of random errors in the data and applicable to large
inbred samples. In addition to relationship misspecifica-
tion, our method can detect some other problems such as
sample duplications and switches.
The method is defined as follows: first, classify each
pair of individuals according to their assumed relation-
ship (such as sib-pairs, parent–offspring pairs, unrelated
individuals, etc.). Second, calculate the mean (µij) and
variance (σij) of identical-by-state allele sharing over a
number of polymorphic loci for each pair of individuals,
i and j. If the sample is homogeneous, we expect each
group to display a characteristic pattern of allele sharing.
For example, sib-pairs will be expected to share more al-
leles on average than unrelated individuals, while parent–
offspring pairs (which share at least one chromosome) are
expected to show less variability in allele sharing than sib-
pairs (which may share zero, one or two chromosomes).
A convenient way to identify individuals with patterns of
allele sharing inconsistent with their specified relationship
is to colour code and plot these mean—variance statistics
The figure presents typical results for a genome scan in
a non-inbred sample. Several distinct clusters are present:
unrelated individuals have the lowest average sharing and
high variance (coloured in blue); half-siblings have higher
sharing on average (coloured in green) and full-siblings
have even higher sharing (coloured in red); parent–
offspring pairs have a similar degree of allele sharing to
sib-pairs but with lower variance (coloured in yellow). All
c ? Oxford University Press 2001
Fig. 1. Sample screen shot. Features described in text.
other relative pairs are grouped together and not displayed
by default. Note that some sibling and full-sibling pairs
have been misclassified and appear in other clusters. A
right corner) and corresponds to a pair of identical twins.
To ensure that outlier points are easily identifiable, GRR
implements an outlier rating scheme and places likely
outliers on top of less interesting points. This scheme
is implemented by calculating the mean and variance
of each allele-sharing statistic within each relationship
group. Then each individual’s scores are standardized
to obtain Zµijand Zσijand assigned the outlier scores
are layered on top of lower rated points. Alternative
schemes for layering data points, such as the Mahalanobis
(1936) distance, can be selected by the user.
GRR recognizes standard genetic formats for genotype
and family structure data, including linkage and QTDT
format files (Ott, 1991; Abecasis et al., 2000). Interactive
features allow the user to select individual families and
inspect statistics for any pair of individuals by clicking the
appropriate plot area.
This approach is simple to implement and can be
incorporated into many genetic analysis databases and
quality control protocols. The method performs efficiently
in genome scan linkage panels, although as few as 50
unlinked markers may be sufficient to verify first-degree
relationships in family samples or to verify that no close
relatives or gross stratification are present in samples of
This research was supported by the Wellcome Trust and
by grant EY-12562 from the National Institutes of Health,
Abecasis,G.R., Cardon,L.R. and Cookson,W.O.C. (2000) A general
test of association for quantitative traits in nuclear families. Am.
J. Hum. Genet., 66, 279–292.
Boehnke,M. and Cox,N.J. (1997) Accurate inference of relation-
ships in sib-pair linkage studies. Am. J. Hum. Genet., 61, 423–
Broman,K.W. and Weber,J.L. (1998) Estimation of pairwise rela-
tionships in the presence of genotyping errors. Am. J. Hum.
Genet., 63, 1563–1564.
Epstein,M.P., Duren,W.L. and Boehnke,M. (2000) Improved infer-
ence of relationship for pairs of individuals. Am. J. Hum. Genet.,
Goring,H.H. and Ott,J. (1997) Relationship estimation in affected
sib pair analysis of late-onset diseases. Eur. J. Hum. Genet., 5,
Mahalanobis,P.C. (1936) On the generalized distance in statistics.
Proc. Natl Inst. Sci. India, 2, 49.
McPeek,M.S. and Sun,L. (2000) Statistical tests for detection of
misspecified relationships by use of genome-screen data. Am. J.
Hum. Genet., 66, 1076–1094.
Ott,J. (1991) Analysis of Human Genetic Linkage. Johns Hopkins
University Press, Baltimore.