> HJGy /bjbjEE 4&''/!1055555lllz1|1|1|1|1|1|1$+35.1lllll15571l^55z1lz15M"f1101555b0lllllll11lll1llll5lllllllll : Text S1
Relationship between European MAF/LD Score and Measure P-Value
Genome wide association studies have more power to detect SNP-trait associations for SNPs that have relatively high MAFs [1]. SNPs that have a high degree of linkage disequilibrium with surrounding SNPs are also more likely to be associated with a trait, as they have an increased likelihood of being in linkage disequilibrium with a causative variant. Whatever the exact cause, we showed in Figure 1 that GWAS SNPs tend to have a higher MAF and a higher LD score in Europe (see Methods) than the average SNP in the Li et al. [2] dataset. Because of this bias, we calculated p-values in our analyses by directly comparing each GWAS SNP to a select subset of SNPs from our full dataset that resembled it in terms of MAF and LD score in Europe. After conducting our p-value calculations, we wanted to ensure that our methods were successful in removing any advantages that might be given to GWAS SNPs in our analysis as a result of their high MAFs and LD scores in Europe. That is, we wanted to test directly whether SNPs with higher MAFs or LD scores tended to have lower p-values for any of our measures. To do this, we calculated the Kendall tau rank correlation coefficient between MAF in Europe and p-value for all GWAS SNPs for each of our 45 measures. We also calculated the p-values of these correlation coefficients. We then repeated these calculations, comparing instead LD score in Europe and measure p-values. The results are shown in Figure S8. Scatter plots for several of these comparisons are shown in Figure S9. For the tests involving European MAF, since we calculated a total of 45 coefficients, the (Bonferroni-corrected) p-value cut-off is 0.001111 for significance at the 0.05 level. Based on this value, there are two cases where p-values for a measure correlate significantly with MAF in Europe: FHG and FAA. In both of these cases, the correlation coefficient is actually positive, which means that a given SNP tends to have a higher p-value if it has a high MAF in Europe. GWAS SNPs are then somewhat disadvantaged for the FHG and FAA measures given their higher MAFs in Europe and our p-values for GWAS SNPs for these two measures are probably somewhat conservative. P-values for iEurope and European MAF also had a positive correlation coefficient with a low p-value, meaning that high MAF SNPs might be disadvantaged for this measure as well. In contrast, the correlation between FWorld p-values and MAF in Europe was negative with a low p-value, so our p-value calculations for this measure may be slightly exaggerated. However, we do not feel as though this possibility significantly alters our results. All other p-values for the 45 tests involving MAF in Europe are larger than 0.01. For the tests comparing LD score to measure p-value, FHG and FAA again had positive correlation coefficients with low p-values. This reinforces our suggestion that our p-values for GWAS SNPs for these two measures may be somewhat conservative. None of the p-values for the remaining tests involving LD score seem particularly notable, given the number of tests we conducted.
Is it possible that the higher MAFs and LD scores of GWAS SNPs in Europe explain our observation that GWAS SNP p-values tend to be lower for measures that involve Eurasian and East Asian populations? We believe this is unlikely given the results shown in Figures S8 and S9. If p-values were randomly assigned to GWAS SNPs, then we would expect about 5% of these SNPs to have p-values of 0.05 or less for each measure. This is about what we observe for measures involving African, Oceanic, and American populations with a few exceptions (see Figure 2 and S2). For many measures involving Eurasian and East Asian populations, more than 5% of GWAS SNPs have p-values of 0.05 or less. This could be explained by the generally elevated MAFs and LD scores of GWAS SNPs in Europe, if MAF and LD score negatively correlated with p-values for these measures. However, as shown in Figures S8 and S9, the correlation coefficients between MAF/LD score and p-value for the majority of these measures do not appear to be significantly different from 0.
References
1. Sale MM, Mychaleckyj JC, Chen W-M (2009) Planning and Executing a Genome Wide Association Study (GWAS). In: Park-Sarge Q-K, Curry TE, editors. Molecular Endocrinology: Methods and Protocols. New York: Humana Press. pp. 403-418.
2. Li JZ, Absher DM, Tang H, Southwick AM, Casto AM, et al. (2008) Worldwide human relationships inferred from genome-wide patterns of variation. Science 2008, 319: 1100-1104.
FGH 3 b l
#
X
Y
9VW
=>@BRmDdսչŵŵť͝ŝhjh*qh8h*khh\khkhxXh1ih-nhh|yhZ*h}]zhmhh7 hZ*5hoho5 ho5 hS5 h"5 h5 he5:Gbtuij !"#$%&'()*+,-gdzmgde&G_`w3Av%5Vh"`c
Iijso"@rstuܼԵ hzm5 h5h7h7hxXh5h3hT[hhhmh*khjhh6666666666666666666666666666666666666666666666666hH6666666666666666666666666666666666666666666666666666666666666666662 0@P`p2( 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p8XV~ OJPJQJ_HmH nH sH tH J`JNormaldCJ_HaJmH sH tH DA D
Default Paragraph FontRiR0Table Normal4
l4a(k (
0No List\\0Subtle Emphasis&6B*PJ\]^JaJmH phsH FFDecimal Aligned
hPJtt
Table Grid7:V0dPK![Content_Types].xmlj0Eжr(Iw},-j4 wP-t#bΙ{UTU^hd}㨫)*1P' ^W0)T9<l#$yi};~@(Hu*Dנz/0ǰ$X3aZ,D0j~3߶b~i>3\`?/[G\!-Rk.sԻ..a濭?PK!֧6_rels/.relsj0}Q%v/C/}(h"O
= C?hv=Ʌ%[xp{۵_Pѣ<1H0ORBdJE4b$q_6LR7`0̞O,En7Lib/SeеPK!kytheme/theme/themeManager.xmlM
@}w7c(EbˮCAǠҟ7՛K
Y,
e.|,H,lxɴIsQ}#Ր ֵ+!,^$j=GW)E+&
8PK!Ptheme/theme/theme1.xmlYOo6w toc'vuر-MniP@I}úama[إ4:lЯGRX^6؊>$!)O^rC$y@/yH*)UDb`}"qۋJחX^)I`nEp)liV[]1M<OP6r=zgbIguSebORD۫qu gZo~ٺlAplxpT0+[}`jzAV2Fi@qv֬5\|ʜ̭NleXdsjcs7f
W+Ն7`gȘJj|h(KD-
dXiJ؇(x$(:;˹!I_TS1?E??ZBΪmU/?~xY'y5g&/ɋ>GMGeD3Vq%'#q$8K)fw9:ĵ
x}rxwr:\TZaG*y8IjbRc|XŻǿI
u3KGnD1NIBs
RuK>V.EL+M2#'fi~Vvl{u8zH
*:(W☕
~JTe\O*tHGHY}KNP*ݾ˦TѼ9/#A7qZ$*c?qUnwN%Oi4=3ڗP
1Pm\\9Mؓ2aD];Yt\[x]}Wr|]g-
eW
)6-rCSj
id DЇAΜIqbJ#x꺃6k#ASh&ʌt(Q%p%m&]caSl=X\P1Mh9MVdDAaVB[݈fJíP|8քAV^f
Hn-"d>znǊ ة>b&2vKyϼD:,AGm\nziÙ.uχYC6OMf3or$5NHT[XF64T,ќM0E)`#5XY`פ;%1U٥m;R>QDDcpU'&LE/pm%]8firS4d7y\`JnίIR3U~7+#mqBiDi*L69mY&iHE=(K&N!V.KeLDĕ{D vEꦚdeNƟe(MN9ߜR6&3(a/DUz<{ˊYȳV)9Z[4^n5!J?Q3eBoCMm<.vpIYfZY_p[=al-Y}Nc͙ŋ4vfavl'SA8|*u{-ߟ0%M07%<ҍPK!
ѐ'theme/theme/_rels/themeManager.xml.relsM
0wooӺ&݈Э5
6?$Q
,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-![Content_Types].xmlPK-!֧6+_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!Ptheme/theme/theme1.xmlPK-!
ѐ' theme/theme/_rels/themeManager.xml.relsPK]
/ &/-/`,#AA@0(
B
S ? tz1F#/13#XX
u.1mllP='yxt
c,U npF6^G*.7ekQm1n>!SSyD6%dO / 1
C
dE}%>
M
7Rv}A}NQ}TRtsvkQ-R"Pf4m .FTA}z)HqbPmy>\MeTgpP`8^z?#@
_S,S_8eFDEvWQ?YwM
,879pv2hJ d < '$!!!/L""!T"J""a"o##gk$;T%h%>%Js%(;%~P&%&p&~p&=Oq'`(F)3*);E))"8),r)|*LF+y+Z:.+-A+)^n++x+I+ ,U-HB.[.r..Rr0r~0009l1
^>|t>y?Y94@*J@k7_@ x@gd@NA&B;BfC4ZCvCCq?CD8;ELPEZrEJEs>F$FeFbdH?%vH HV
Hv[IIpsJFJC2&KK"PK@ LRSLzLjL4L4OM2xMcMh*NPOHOO&HO?O:iPsP9PPlPf)QuQnvQB
PRSURt2SY`StnS{&T)g9T7@"sz+s#ds`s t^Vt(et u5N+u[Du*v&v|vuwwTwOExxxzy)yhy\syCy{yl^z|@z.yzu#3{~p?{\P{O|=|_}}}SD}}0~qRpF;PQSdjk8vE@ABCDEFIRoot Entry FMKData
1Table5WordDocument4&SummaryInformation(7DocumentSummaryInformation8?CompObjy
F'Microsoft Office Word 97-2003 Document
MSWordDocWord.Document.89q