文章的題目是A bidirectional Mendelian randomization study supports causal effects of kidney function on blood pressure，這篇文章用到的統(tǒng)計(jì)技巧叫做Two-sample MR----兩樣本孟德爾隨機(jī)分析。我還查閱了別的孟德爾隨機(jī)化的文獻(xiàn)，這個(gè)Two-sample MR的分析其實(shí)是非常常用的。
Two-sample MR分析的一般步驟
第一步是找工具變量，我們要的是基因作為工具變量這些個(gè)基因都是從別人的研究中挑出來的，所有的基因研究有個(gè)專門的庫叫做genome wide association studies (GWAS)。我們需要做的就是從這個(gè)庫中挑出來我們自己需要的和我們暴露相關(guān)的基因變量SNPs。這是第一步。第二步就是估計(jì)我們的工具變量對(duì)結(jié)局的作用，工具變量對(duì)結(jié)局的作用也是從所有的研究中估計(jì)出來的整體效應(yīng)，這樣可以拒絕單個(gè)研究的偏倚。第三步就是合并多個(gè)SNP的效應(yīng)量，這個(gè)效應(yīng)量是我們得到暴露和結(jié)局因果效應(yīng)的前提。第四步就是用合并后的數(shù)據(jù)進(jìn)行孟德爾隨機(jī)化分析和相應(yīng)的敏感性分析。做分析的整個(gè)流程就在下面的圖中啦：
To ensure independence among genetic instruments, we applied LD clumping60 with a clumping window of 10 MB and an r2 cutoff of 0.001 (default of the ld_clumpfunction)
作者有把篩出來的變量的暴露和結(jié)局的工具變量展示在文獻(xiàn)中（但是放在補(bǔ)充材料中的，我并沒有能找到，欸）然后就到第二步和第三步，估計(jì)工具變量對(duì)暴露和結(jié)局的作用，這個(gè)時(shí)候要考慮工具變量一定不能直接影響結(jié)局（叫做pleiotropy），所以作者會(huì)用好幾個(gè)算法（inverse variance weighted method, mendelian randomisation-Egger (MR-Egger) method, weighted median method, and weighted mode based estimation）來估計(jì)SNP的作用，并將多個(gè)SNP的效應(yīng)合并，用到的是harmonise_data這個(gè)函數(shù)。最后得到一個(gè)總的合并后的效應(yīng)，這個(gè)效應(yīng)被認(rèn)為是比較穩(wěn)健的：
We applied four complementary methods of two sample mendelian randomisation (inverse variance weighted method, mendelian randomisation-Egger (MR-Egger) method, weighted median method, and weighted mode based estimation), which make different assumptions about horizontal pleiotropy. A consistent effect across the four methods is less likely to be a false positive.
Two sample Mendelian randomisation (2SMR) is a method to estimate the causal effect of an exposure on an outcome using only summary statistics from genome wide association studies (GWAS).
具體地，作者使用的R包是'TwoSampleMR’。使用TwoSampleMR的基本流程包括4步：
選擇工具變量
從GWAS數(shù)據(jù)庫提取工具變量
合并效應(yīng)量
做MR分析，敏感性分析，畫圖，出報(bào)告
在分析中我們要使用SNPs來作為工具變量：
Mendelian randomization is a method to assess the causal effect of an exposure on an outcome using an instrument, defined by one or more SNPs, as a proxy for the exposure.
Methodological advances mean that Mendelian randomization can be implemented using summary statistics from GWAS, without individual level data. This requires SNP-exposure associations and SNP-outcome associations obtained from separate datasets and is known as two-sample Mendelian randomization
這里又有許多術(shù)語需要給大家做做鋪墊：
孟德爾隨機(jī)化：Mendelian randomization is a method to assess the causal effect of an exposure on an outcome using an instrument, defined by one or more single nucleotide polymorphisms, as a proxy for exposure.（SNP就是工具變量）
Genome-wide association study (GWAS)：Genome-wide association studies identify the genetic variants that are associated with a given phenotype.和暴露（表型）相關(guān)的基因都是從GWAS中找的
Heterogeneity：Heterogeneity is defined as the variation in the causal estimate across SNPs.這個(gè)是不同SNP效應(yīng)的異質(zhì)性，所以多個(gè)SNP的效應(yīng)是需要合并的
同時(shí)因?yàn)槲覀兪沁x了很多個(gè)SNP進(jìn)行的效應(yīng)估計(jì)，我們可以出整個(gè)SNP效應(yīng)的森林圖和漏斗圖，漏斗圖可以幫我們看SNP的異質(zhì)性，我們還可以用Cochran’s Q and I2統(tǒng)計(jì)量檢驗(yàn)不同SNP的異質(zhì)性，然后選擇正確的方法對(duì)各個(gè)SNP的效應(yīng)進(jìn)行合并得到總效應(yīng)：
