> q` bjbjqPqP .::
Jprrrrrrr,......$'h
RRErrrrrRrrr:rr,r,rV4,0
rrrrrrrRRrrrrrrr
Text S1
Let A and B be alleles at two bi-allelic polymorphisms (e.g. SNPs) located close enough to be in LD. The standard LD coefficient (DAB) is transformed to the LD correlation coefficient (r) by the well known formula EMBED Equation.3 where PA and PB are allele frequencies of A and B [42]. Similarly, Weir [42,43] described the composite disequilibrium coefficient (DAB) which he explained is similarly transformed to an analogous correlation coefficient (r*) by the formula EMBED Equation.3 where DA and DB are Hardy-Weinberg disequilibrium coefficients. Whether the formula variables are for the entire population or a particular sample, Zaykin [44] and Weir [43] each noted that r* is the same as the formula for the (unsquared) correlation between genotypes ( EMBED Equation.3 ) at two loci; and thus EMBED Equation.3 .
For a population in Hardy-Weinberg equilibrium (HWE), the population values of the four disequilibrium coefficients are such that DAB=DAB and DA=DB=0; and Weir [43] also noted that the population values of r and r* become identical implying that EMBED Equation.3 . Weir [43] also stated that the sample statistics of r and r* are normally distributed after Fisher s z-transforma
&(ĲĄwoo]#jx+M
hfLhNCJPJUVhfLhNH*jhfLhNEHU#jwo+M
hfLhNCJPJUVhfLhNOJQJmH sH jhfLhNEHU#jzl+M
hfLhNCJPJUVhfLhNjhfLhNUhfLhNH*mH sH hfLhNmH sH hNmH sH hnhNh> dh`gdNdhgdNgdNdhgdN
"ದtr`SjChfLhNEHU#j!,M
hfLhNCJPJUVUj hfLhNEHU#j_1,M
hfLhNCJPJUVhfLhNH*hfLhNmH sH hfLhNH*mH sH hfLhNOJQJmH sH jhfLhNEHU#j3x+M
hfLhNCJPJUVhfLhNjhfLhNUjhfLhNEHU tion such that their sample means are equal under HWE while their sampling variances are 1/(2N-3) and 1/(N-3), respectively (where N is number of subjects in the sample). We therefore conclude that EMBED Equation.3 for samples from populations in HWE for which sample size is sufficiently large (N>100) for accurate normal distribution under Fisher z-transformation with the approximation being very good for sample sizes as large as our GWAS (N>1000).
hNhNmH sH hfLhNmH sH hfLhN,1h. A!"#$%Dd
Wb
c$A??3"`?2kTI%)D`!kTI%)hxڝKP]ئU"Dju
N"mW
֊J"8 PQ\
wb}s=i$`IrKc!cc&٠eb][v7f$s{Z,g&rlBG;('-d6)^i]X,rYUņ.A7^0R8kDT~("<Dd
b
c$A??3"`?2={mc@y#5Z+`!{mc@y#5Z+&,hxڝ1OQ{wX)9)(HL1Al" \b"Y_F7ގy 4773E\
Pz)bfRR)[ՄDWxhhTD_?%)rFM
?T1Vlv|:S=gS8?8b"n^]68Dd
b
c$A??3"`?2I~[rLcp73n%`!~[rLcp73n`:hxcdd``> @c112BYL%bpu41b;.+KRsA<.'f8lk4Dd
b
c$A??3"`?2~-;7Kv5uZ`!R-;7Kv5ud` h xڝQJ@fIx)C¢i->oazГ:
.;7220+ۧ:iy]Y.n-hCF"Y4ߩ9]j~TpgD3U&*^Go/br X;_jpAGd\QgS${/[{:aǟ]qe'*?/nPK o{Pq=e9+=)NLUnѬ'9e
!#"$&%'()*+,-./0Root Entry# FngData
WordDocument".ObjectPool% Vng_1294691450FVVOle
CompObjfObjInfo
!$')*+,-.012356789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_a
FMicrosoft Equation 3.0DS EquationEquation.39q
r=)DAB
PA
(1"PA
)PB
(1"PB
)Equation Native _1294692215FVVOle
CompObj
f
FMicrosoft Equation 3.0DS EquationEquation.39q
r*=)AB
[PA
(1"PA
)+DA
] [PB
(1"PB
)+DB
]ObjInfo
Equation Native _1294694643FVVOle
FMicrosoft Equation 3.0DS EquationEquation.39q')d
Rgeno
FMicrosoft Equation 3.0DS EqCompObjfObjInfoEquation Native C_1294694451FVVOle
CompObjfObjInfoEquation Native iuationEquation.39qM T
Rgeno2
=(r*)2
FMicrosoft Equation 3.0DS EquationEquation.39q_1294741855FVVOle
CompObjfObjInfoEquation Native _1294737842FVVOle
"CompObj #fnc
(r*)2
=Rgeno2
=r2
FMicrosoft Equation 3.0DS EquationEquation.39qnA T
Rgeno2
H"r2ObjInfo!%Equation Native &]1Table4
SummaryInformation($(Oh+'0p
,8
DPX`hAppendixrm2Normalacollings2Microsoft Office Word@F#@r@r՜.+,0hDocumentSummaryInformation8/@CompObj`qp
$The Wellcome Trust Sanger Institute AppendixTitleN@NNNormal$1$a$CJKH_HmH nHsH tH@@@N Heading 2$@&OJPJQJDADDefault Paragraph FontViVTable Normal :V44
la(k(No List _`000000
,.FZ\Vjl::::::JNFIQV333 >N@Բn55@`@`
`@``UnknownGz Times New Roman5Symbol3&z ArialG
MS Mincho-3 fgK1
MS Gothic-3 0000"qh$f$f!242qHP)?N2Appendixrm2 acollings
FMicrosoft Office Word Document
MSWordDocWord.Document.89q