利用FUSION方法计算可变剪接sQTL TWAS模型

1 利用FUSION方法计算可变剪接sQTL TWAS模型

表型数据为leafcutter软件计算的内含子切除率数据，表型格式为chr1_12227_12595_clu_1_+。

2 运行命令

直接运行一下命令

bash 复制代码

target_plink=/.../plink
phenotype_file=/.../AS/AS.txt
coordinate_file=/.../AS_coordinate/AS_coordinate.txt
fusion_ldref=/.../LD_reference/GRCh38/EAS/LDREF
fusion_software=/.../fusion_twas
plink=/.../plink
gemma=/.../gemma
gcta=/.../fusion_twas/gcta_nr_robust
output=/.../model_sqtl/
covar=/.../cov.txt

# 第一步构建模型，生成wgt.RDat文件
feature_name="NA"
Rscript ./TWAS_weights_AS.R \
	--PLINK_prefix ${target_plink} \
	--phenotype_file ${phenotype_file} \
	--coordinate_file ${coordinate_file} \
	--gene_name ${feature_name} \
	--plink ${plink} \
	--gcta ${gcta} \
	--gemma ${gemma} \
	--ld_ref_dir ${fusion_ldref} \
	--fusion_software ${fusion_software} \
	--output_dir ${output} \
	--covar ${covar}

Rscript ./OP_packaging_fusion_weights.R \
	--RDat_dir ${output}/Output \
	--coordinate_file ${coordinate_file} \
	--output_name TWAS_Weights \
	--output_dir ${output}/TWAS_weights_package

3 输入文件详情

3.1 plink 为vcf数据，plink格式文件

3.2 AS.txt为内含子剪接率矩阵

格式如下

bash 复制代码

head -5 AS.txt | awk '{print $1, $2, $3, $4, $5}'
FID IID chr1_12227_12595_clu_1_+ chr1_12227_12613_clu_1_+ chr1_12227_12646_clu_1_+
A1 A1 -1.117206e+00 1.173844e+00 3.630778e-01
A2 A2 8.535833e-01 -4.706559e-01 5.331248e-01
A3 A3 8.869868e-01 -8.779329e-01 -3.090774e-01
A4 A4 0.01870461 -0.12663100 0.01870461

3.3 AS_coordinate.txt文件为内含子位置信息

格式如下

bash 复制代码

head AS_coordinate.txt
Chr	start	end	ID
1	12227	12595	chr1_12227_12595_clu_1_+
1	12227	12613	chr1_12227_12613_clu_1_+
1	12227	12646	chr1_12227_12646_clu_1_+
1	12697	13221	chr1_12697_13221_clu_2_+
1	12697	13403	chr1_12697_13403_clu_2_+
1	12697	13453	chr1_12697_13453_clu_2_+
1	12697	13483	chr1_12697_13483_clu_2_+
1	12721	13221	chr1_12721_13221_clu_2_+
1	12721	13403	chr1_12721_13403_clu_2_+```

3.4 /.../LD_reference/GRCh38/EAS/LDREF为LD reference所在文件夹

bash 复制代码

1000G.EAS.10.bed    1000G.EAS.12.map    1000G.EAS.15.bim    1000G.EAS.17.nosex  1000G.EAS.1.fam     1000G.EAS.21.ped    1000G.EAS.3.log    1000G.EAS.6.bed    1000G.EAS.8.map
1000G.EAS.10.bim    1000G.EAS.12.nosex  1000G.EAS.15.fam    1000G.EAS.17.ped    1000G.EAS.1.log     1000G.EAS.22.bed    1000G.EAS.3.map    1000G.EAS.6.bim    1000G.EAS.8.nosex
1000G.EAS.10.fam    1000G.EAS.12.ped    1000G.EAS.15.log    1000G.EAS.18.bed    1000G.EAS.1.map     1000G.EAS.22.bim    1000G.EAS.3.nosex  1000G.EAS.6.fam    1000G.EAS.8.ped
1000G.EAS.10.log    1000G.EAS.13.bed    1000G.EAS.15.map    1000G.EAS.18.bim    1000G.EAS.1.nosex   1000G.EAS.22.fam    1000G.EAS.3.ped    1000G.EAS.6.log    1000G.EAS.9.bed
1000G.EAS.10.map    1000G.EAS.13.bim    1000G.EAS.15.nosex  1000G.EAS.18.fam    1000G.EAS.1.ped     1000G.EAS.22.log    1000G.EAS.4.bed    1000G.EAS.6.map    1000G.EAS.9.bim
1000G.EAS.10.nosex  1000G.EAS.13.fam    1000G.EAS.15.ped    1000G.EAS.18.log    1000G.EAS.20.bed    1000G.EAS.22.map    1000G.EAS.4.bim    1000G.EAS.6.nosex  1000G.EAS.9.fam
1000G.EAS.10.ped    1000G.EAS.13.log    1000G.EAS.16.bed    1000G.EAS.18.map    1000G.EAS.20.bim    1000G.EAS.22.nosex  1000G.EAS.4.fam    1000G.EAS.6.ped    1000G.EAS.9.log
1000G.EAS.11.bed    1000G.EAS.13.map    1000G.EAS.16.bim    1000G.EAS.18.nosex  1000G.EAS.20.fam    1000G.EAS.22.ped    1000G.EAS.4.log    1000G.EAS.7.bed    1000G.EAS.9.map
1000G.EAS.11.bim    1000G.EAS.13.nosex  1000G.EAS.16.fam    1000G.EAS.18.ped    1000G.EAS.20.log    1000G.EAS.2.bed     1000G.EAS.4.map    1000G.EAS.7.bim    1000G.EAS.9.nosex
1000G.EAS.11.fam    1000G.EAS.13.ped    1000G.EAS.16.log    1000G.EAS.19.bed    1000G.EAS.20.map    1000G.EAS.2.bim     1000G.EAS.4.nosex  1000G.EAS.7.fam    1000G.EAS.9.ped
1000G.EAS.11.log    1000G.EAS.14.bed    1000G.EAS.16.map    1000G.EAS.19.bim    1000G.EAS.20.nosex  1000G.EAS.2.fam     1000G.EAS.4.ped    1000G.EAS.7.log    
1000G.EAS.11.map    1000G.EAS.14.bim    1000G.EAS.16.nosex  1000G.EAS.19.fam    1000G.EAS.20.ped    1000G.EAS.2.log     1000G.EAS.5.bed    1000G.EAS.7.map    
1000G.EAS.11.nosex  1000G.EAS.14.fam    1000G.EAS.16.ped    1000G.EAS.19.log    1000G.EAS.21.bed    1000G.EAS.2.map     1000G.EAS.5.bim    1000G.EAS.7.nosex  
1000G.EAS.11.ped    1000G.EAS.14.log    1000G.EAS.17.bed    1000G.EAS.19.map    1000G.EAS.21.bim    1000G.EAS.2.nosex   1000G.EAS.5.fam    1000G.EAS.7.ped    
1000G.EAS.12.bed    1000G.EAS.14.map    1000G.EAS.17.bim    1000G.EAS.19.nosex  1000G.EAS.21.fam    1000G.EAS.2.ped     1000G.EAS.5.log    1000G.EAS.8.bed    
1000G.EAS.12.bim    1000G.EAS.14.nosex  1000G.EAS.17.fam    1000G.EAS.19.ped    1000G.EAS.21.log    1000G.EAS.3.bed     1000G.EAS.5.map    1000G.EAS.8.bim    
1000G.EAS.12.fam    1000G.EAS.14.ped    1000G.EAS.17.log    1000G.EAS.1.bed     1000G.EAS.21.map    1000G.EAS.3.bim     1000G.EAS.5.nosex  1000G.EAS.8.fam    
1000G.EAS.12.log    1000G.EAS.15.bed    1000G.EAS.17.map    1000G.EAS.1.bim     1000G.EAS.21.nosex  1000G.EAS.3.fam     1000G.EAS.5.ped    1000G.EAS.8.log

3.5 fusion_twas为fusion安装包所在的位置

3.6 plink、gemma、gcta_nr_robust分别是各软件位置

4 所需要的R程序

4.1 TWAS_weights_AS.R

该文件在资源里寻找

注意：TWAS_weights_AS.R里需要修改t.awk的路径

bash 复制代码

cat t.awk
NR==1 {
    for (i=1; i<=NF; i++) {
        ix[$i] = i
    }
}
NR>1 {
    print $ix[c1], $ix[c2], $ix[c3]
}

4.2 TWAS_weights_AS.R里调用了FUSION.compute_weights.R程序

该文件在资源里寻找

参考：

leafcutter : https://davidaknowles.github.io/leafcutter/

FUSION: http://gusevlab.org/projects/fusion/