SAS/Genetics Software
Enterprise Intelligence | Supplier Intelligence | Organizational Intelligence | Customer Intelligence | Intelligence Architecture
SAS/Genetics™ Software
Performing genetics research with
dedicated analysis tools
As a statistical geneticist or biostatisti-
cian, you may struggle with adequate
tools to help you perform your research.
Available analysis software may be useful,
but you can waste time customizing it to
perform specific tasks. The results may
Most importantly, SAS/Genetics comes
be disappointing and it may not provide
from SAS. The world’s leading pharma-
adequate support when you need help.
ceutical and biotechnology organizations
An ideal solution would handle your
have chosen SAS as their strategic
specialized research needs – efficiently
partner – trusting us to help them use
and easily – and come with seasoned
technology to know more about their
experts who could respond to your
data and their processes and to better
questions on a 24/7 basis.
anticipate results. We’ve been in the
business of knowledge discovery for
SAS/Genetics™ software is designed
over a quarter-century and we’ll be here
specifically for genetics research.
over the life of your investment.
It enables you to study genetic markers
individually, examine the relationship
between multiple markers and find associ-
ations between markers and traits using
case-control or family data. And all your
results can be depicted graphically – so
they’re easily understandable and share-
able with others–with output from numerous
procedures interpreted simultaneously.
A powerful new
product among SAS’
scientific discovery solutions
SAS/Genetics™ contains five statistical procedures (or PROCs) and one macro. Based on SAS statistical methods, these PROCs
are accessed through base SAS software.
This procedure provides summary statistics to characterize individual markers. PROC
ALLELE calculates measures of marker informativeness, which can help determine which
markers are most valuable for association or linkage studies. You can produce tables of
PROC ALLELE
allele and genotype frequencies, including bootstrap confidence intervals for hypothesis
testing. An asymptotic and exact test can be used for testing markers for Hardy-Weinberg
equilibrium (HWE).
PROC ALLELE also explores associations between two markers or pairs of alleles at two
markers. It can calculate:
• Linkage disequilibrium between pairs of alleles using haplotypes entered by the user.
• Maximum likelihood estimates of the two-locus haplotype frequencies.
• Composite linkage disequilibrium coefficients that do not require haplotype frequencies.
Additionally, you can perform an asymptotic and exact version of an overall test for no
linkage disequilibrium between two markers with any number of alleles.
Haplotypes are sets of specific alleles that occur together on the same chromosome,
PROC HAPLOTYPE
as opposed to genotypes which occur in pairs arising from different chromosomes.
This procedure estimates haplotype frequencies for more than two loci using genotype
data to reconstruct possible haplotypes. Assuming HWE, the expectation-maximization
(EM) algorithm is used to calculate the maximum likelihood estimates of the haplotype
frequencies. Various standard error estimation methods are available, as well as predicted
probabilities of individual haplotypes. These estimated haplotype frequencies can then
be used in tests for linkage disequilibrium across markers or for association with a
r e s
binary variable such as disease status (case-control data).
e
These procedures test markers for association with a dichotomous trait such as disease
PROC CASECONTROL
status. PROC CASECONTROL uses random samples of unrelated individuals, while
and PROC FAMILY
PROC FAMILY analyzes family data consisting of either parents and an affected child, or
sibling pairs made up of one affected and one unaffected sibling.
• Several case-control tests can be used: the genotype case-control test, the allele
case-control test, and the linear trend test with or without a variance inflation correction.
• Family-based tests include the original transmission/disequilibrium test (TDT), the sibling
TDT (S-TDT), the sibling disequilibrium test (SDT) and the reconstruction-combined TDT
(RD-TDT). All tests can handle markers with any number of alleles.
Marker-association tests performed by PROC CASECONTROL and PROC FAMILY
PROC PSMOOTH
examine one marker at a time, but linkage disequilibrium between markers that are close
together causes correlations between these tests. PROC PSMOOTH accounts for this
correlation by smoothing the p-values, where smoothing involves using a function of
p-values from consecutive markers to form a new p-value. PROC PSMOOTH can be
applied directly to output data sets from PROC CASECONTROL and/or PROC FAMILY,
or to user-created data sets containing p-values from linkage or association analyzes.
It can also adjust p-values for multiple testing.
Included in the SAS autocall library, this macro creates a triangular plot for graphically
%TPLOT macro
displaying marker test results in an easy, interpretable manner.
• Output from PROC ALLELE can be combined into this single plot with output from
PROC CASECONTROL, PROC FAMILY, PROC PSMOOTH, or a user-created data set
containing p-values from marker tests.
• You can also represent p-values from the HWE test, tests for linkage disequilibrium
tests between pairs of markers, and linkage or association tests between each
marker and the trait of interest, all smoothed and adjusted using PROC PSMOOTH.
a r c h
From SAS, an industry standard
SAS is committed to helping customers extract from their data the knowledge they need
to make intelligent decisions. Toward that end, we incorporate award-winning training and
technical support as part of every solution. SAS-certified training professionals provide
expert knowledge tailored to meet your needs, using your own data in your work
environment. Additionally, advanced training is available at your site or through our
regional training centers.
SAS is the world’s largest privately held software company, serving more than 37,000
customer sites in 110 countries. More than 3.5 million users have seen the power of SAS
products turn raw data into rare insights.
For more information about SAS/Genetics from SAS visit us on the Web at:
www.sas.com/industry/pharma/index.html
48557US.0302
