Power_Logistic(SKAT)
Power_Logistic()所属R语言包：SKAT
Power calculation, Dichotomous traits
功率计算，二分特征
译者：生物统计家园网 机器人LoveR
描述----------Description----------
Compute an average power of SKAT for testing association between a genomic region and dichotomous phenotypes from case-control studies with a given disease model.
计算的平均功率SKAT测试协会之间的基因组区域和二分表型的病例对照研究特定疾病模型。
用法----------Usage----------
Power_Logistic(Haplotypes = NULL, SNP.Location = NULL, SubRegion.Length=-1
, Prevalence=0.01, Case.Prop=0.5, Causal.Percent=5, Causal.MAF.Cutoff=0.03
, alpha =c(0.01,10^(-3),10^(-6)), N.Sample.ALL = 500 * (1:10)
, Weight.Param=c(1,25), N.Sim=100, OR.Type = "Log"
, MaxOR=5, Negative.Percent=0)
Power_Logistic_R(Haplotypes = NULL, SNP.Location = NULL, SubRegion.Length=-1
, Prevalence=0.01, Case.Prop=0.5, Causal.Percent=5, Causal.MAF.Cutoff=0.03
, alpha =c(0.01,10^(-3),10^(-6)), N.Sample.ALL = 500 * (1:10)
, Weight.Param=c(1,25), N.Sim=100, OR.Type = "Log"
, MaxOR=5, Negative.Percent=0, r.corr=0)
参数----------Arguments----------
参数：Haplotypes
a haplotype matrix with each row as a different individual and each column as a separate SNP (default= NULL). Each element of the matrix should be either 0 (major allel) or 1 (minor allele). If it has NULL, SKAT.haplotype dataset will be used to compute power.
一个单倍型矩阵，不同的个人和每一行，每一列作为一个单独的SNP（默认值= NULL）。矩阵的每个元素应该是0（主要等位基因）或1（次要等位基因）。如果它具有NULL，SKAT.haplotype数据集将被用于计算功率。
参数：SNP.Location
a numeric vector of SNP locations which should be matched with the SNPs in the Haplotype matrix (default= NULL). It is used to obtain subregions. When Haplotype=NULL, it should be NULL.
一个数值向量的SNP的位置应与单核苷酸多态性的单倍型矩阵（默认值= NULL）。它是用来获得分区域。当单倍型= NULL，它应该是NULL。
参数：SubRegion.Length
a value of the length of subregions (default= -1). Each subregion will be randomly selected, and then the average power will be calculated by taking the mean over the estimated powers of all subregions. If SubRegion.Length=-1 (default), the length of the subregion is the same as the length of the whole region, and thus there is no random selection of subregions.
分区域的长度的值（默认值= -1）。各次区域将被随机选中，然后将计算的平均功耗的估计权力的所有次区域的平均。如果SubRegion.Length = -1（默认），该次区域的长度是整个区域的长度一样，因此没有随机选择区域。
参数：Prevalence
a value of disease prevalence.
疾病的患病率值。
参数：Case.Prop
a value of the proportion of case samples. For example, Case.Prop=0.5 means 50 % of samples are cases and 50 % of samples are controls.
的比例的情况下样品的值。例如，Case.Prop = 0.5意味着50％的样品的情况，和50％的样品的控制。
参数：Causal.Percent
a value of the percentage of causal SNPs among rare SNPs (MAF < Causal.MAF.Cutoff)(default= 5).
罕见SNPs之间的因果SNP位点的百分比值（MAF <Causal.MAF.Cutoff）（默认值= 5）。
参数：Causal.MAF.Cutoff
a value of MAF cutoff for the causal SNPs. Only SNPs that have MAFs smaller than this are considered as causal SNPs (default= 0.03).
一个的MAF截止值的因果单核苷酸多态性。只有MAFS比这更小的单核苷酸多态性被认为是因果单核苷酸多态性（默认值= 0.03）。
参数：alpha
a vector of the significance levels (default= c(0.01,10^(-3),10^(-6))).
一个向量的显着性水平（默认目录为C（0.01,10 ^（-3），10 ^（-6）））。
参数：N.Sample.ALL
a vector of the sample sizes (default= 500 * (1:10)).
一个向量的样本大小（默认值= 500 *（1:10））。
参数：Weight.Param
a vector of parameters of beta weights (default= c(1,25)).
一个向量的二级权重的参数（默认值= C（1,25））。
参数：N.Sim
a value of number of causal SNP/SubRegion sets to be generated to compute the average power (default= 100). Power will be computed for each causal SNP/SubRegion set, and then the average power will be obtained by taking mean of the computed powers.
将产生数的因果SNP /次区域设置的值来计算的平均功耗（默认值= 100）。 Power将被计算为每个因果SNP /分区域设置，然后，将通过以下方式获得的平均功率平均值计算的权力。
参数：OR.Type
a function type of effect sizes (default= “Log”). “Log” indicates that log odds ratio of each causal variant equals to c|log10(MAF)|, and “Fixed” indicates that log odds ratio of all causal variants are the same.
一个功能型的影响的大小（默认值=“log”）。 “log”表示每个因果变种该log的比值比等于c|log10(MAF)|，和“固定”表示该log的比值比，所有的因果变化是相同的。
参数：MaxOR
a numeric value of the maximum odds ratio (default= 5). When OR.Type=“Log”, the maximum odds ratio is MaxOR (when MAF=0.0001). When OR.Type=“Fixed”, all causal variants have the same odds ratio (= MaxOR). See details
一个数值的最大比值比（默认值= 5）。当OR.Type =“登录”，是MaxOR的最大比值比（MAF = 0.0001）。当OR.Type =“固定”，所有的因果变化具有相同的比值比（= MaxOR）。查看详细资料
参数：Negative.Percent
a numeric value of the percentage of coefficients of causal variants that are negative (default= 0).
一个数值的因果变化的系数是负的（默认值= 0）的百分比。
参数：r.corr
(Power_Logistic_R only) the ρ parameter of new class of kernels with compound symmetric correlation structure for genotype effects (default= 0). See details.
（Power_Logistic_R）ρ参数的一类新的复合对称的相关结构的内核基因型效应（默认值= 0）。查看详细信息。
Details
详细信息----------Details----------
By default it use the haplotype information in the SKAT.haplotypes dataset. So you can left Haplotypes and SNP.Location as NULL if you want to use the SKAT.haplotypes dataset.
默认情况下，使用单倍型信息中的SKAT.haplotypes的数据集。所以，你可以为NULL，如果你要使用的SKAT.haplotypes数据集的单倍型和SNP.Location。
When OR.Type="Log", MaxOR is the odd ratio of the causal SNP with MAF = 10^{-4} and used to obtain c value in the function \log OR = c|log10(MAF)|. For example, if MaxOR=5, c = log(5)/4 = 0.402. Then a variant with MAF=0.001 has log odds ratio = 1.206 and a variant with MAF=0.01 has log odds ratio = 0.804.
当OR.Type =“登录”，MaxOR的因果SNP的MAF是奇比= 10^{-4}和使用的功能\log OR = c|log10(MAF)|c值。例如，如果MaxOR = 5，c = log(5)/4 = 0.402。与MAF的一个变种log的比值比= 0.001 = 1.206和一个变种，MAFlog的比值比= 0.01 = 0.804。
When the SubRegion.Length is small such as 3kb or 5kb, it is possible that you can have different estimated power for each run with N.Sim = 50 \sim 100. Then, please increase N.Sim to 500 \sim 1000 to obtain stable results.
3KB或5KB如当的SubRegion.Length是小，它是可能的，你可以有不同的估计功率为每个运行与N.Sim =50 \sim 100。然后，请增加N.Sim，500 \sim 1000获得稳定的结果。
Power_Logistic_R computes the power with new class of kernels with compound symmetric correlation structure. It uses a slightly different method to compute power, and thus Power_Logistic and Power_Logistic_R will produce slightly different results although r.corr=0.
Power_Logistic_R计算功率与类新的复合对称的相关结构的内核。它采用了稍微不同的方法来计算电源，从而Power_Logistic Power_Logistic_R的会产生稍微不同的结果，虽然r.corr = 0。
值----------Value----------
参数：Power
A matrix with each row as a different sample size and each column as a different significance level. Each element of the matrix is the estimated power.
A矩阵为不同的样本大小，每行和每列不同的显着性水平。该矩阵的每个元素是估计功率。
（作者）----------Author(s)----------
Seunggeun Lee
实例----------Examples----------
#[]
# Calculate the average power of randomly selected 3kb regions [随机选择的3kb的区域计算的平均功率]
# with the following conditions.[满足下列条件。]
#[]
# Causal percent = 20%[因果％= 20％]
# Negative percent = 20%[负百分比= 20％]
# Max OR = 7 at MAF = 10^-4[最大OR = 7 MAF = 10 ^ -4]
#[]
# When you use this function, please increase N.Sim (more than 100) [当您使用此功能，请增加N.Sim（超过100个）]
#[]
out.b<-Power_Logistic(SubRegion.Length=3000,
Causal.Percent= 20, N.Sim=10 ,MaxOR=7,Negative.Percent=20)
out.b
#[]
# Calculate the required sample sizes to achieve 80% power[计算所需的样本量达到80％的电力]
Get_RequiredSampleSize(out.b, Power=0.8)
转载请注明:出自 生物统计家园网(http://www.biostatistic.net)。
注：
注1：为了方便大家学习，本文档为生物统计家园网机器人LoveR翻译而成，仅供个人R语言学习参考使用，生物统计家园保留版权。
注2：由于是机器人自动翻译，难免有不准确之处，使用时仔细对照中、英文内容进行反复理解，可以帮助R语言的学习。
注3：如遇到不准确之处，请在本贴的后面进行回帖，我们会逐渐进行修订。
|
发表于 2012-9-30 09:49:39
