-
Notifications
You must be signed in to change notification settings - Fork 3
/
analysis_plan.Rmd
260 lines (172 loc) · 6.92 KB
/
analysis_plan.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
---
title: "QGT-Columbia-analysis-plan"
author: "Hae Kyung Im"
date: "2020-06-03"
output: workflowr::wflow_html
editor_options:
chunk_output_type: console
---
```{r preliminary definitions}
library(tidyverse)
```
# Preliminary information
Data and copies of repositories can be downloaded from [Box here](https://uchicago.box.com/s/zhapf2zfxcpj7thvq4sjnqale3emleum)
The latest version of the analysis plan that generated this page is on [github here](https://github.com/hakyimlab/QGT-Columbia-HKI/blob/master/analysis/analysis_plan.Rmd)
# Transcriptome-wide association methods
```{r preliminaries}
print(getwd())
pre="/home/rstudio/QGT-Columbia-HKI"
model.dir=glue::glue("{pre}/models")
metaxcan.dir=glue::glue("{pre}/repos/MetaXcan-master/software")
fastenloc.dir=glue::glue("{pre}/repos/fastenloc-master")
torus.dir=glue::glue("{pre}/repos/torus-master")
twmr.dir=glue::glue("{pre}/repos/TWMR-master")
results.dir=glue::glue("{pre}/results")
```
## predict expression
```{bash predict genetic component of expression,eval=FALSE}
export PRE="/home/rstudio/QGT-Columbia-HKI"
export DATA=$PRE/predixcan/data
export MODEL=$PRE/models
export METAXCAN=$PRE/repos/MetaXcan-master/software
export RESULTS=$PRE/results
printf "Predict expression\n\n"
python3 $METAXCAN/Predict.py \
--model_db_path $PRE/models/gtex_v8_en/en_Whole_Blood.db \
--vcf_genotypes $DATA/genotype/ALL.chr22.shapeit2_integrated_snvindels_v2a_27022019.GRCh38.phased.vcf.gz \
--vcf_mode genotyped \
--variant_mapping $DATA/gtex_v8_eur_filtered_maf0.01_monoallelic_variants.txt.gz id rsid \
--on_the_fly_mapping METADATA "chr{}_{}_{}_{}_b38" \
--prediction_output $RESULTS/predixcan/Whole_Blood__predict.txt \
--prediction_summary_output $RESULTS/predixcan/Whole_Blood__summary.txt \
--verbosity 9 \
--throw
```
## assess prediction performance (optional)
```{r}
predicted_expression = read_tsv(glue::glue("{results.dir}/predixcan/Whole_Blood__predict.txt"))
dim(predicted_expression)
head(predicted_expression[,1:5])
prediction_summary = read_tsv(glue::glue("{results.dir}/predixcan/Whole_Blood__summary.txt"))
dim(prediction_summary)
head(prediction_summary)
## merge with GEUVADIS expression data
## calculate spearman correlation
## select a few genes and plot predicted vs observed expression
```
## run association with phenotype
```{bash, eval=FALSE}
export PRE="/home/rstudio/QGT-Columbia-HKI"
export DATA=$PRE/predixcan/data
export MODEL=$PRE/models
export METAXCAN=$PRE/repos/MetaXcan-master/software
export RESULTS=$PRE/results
printf "association\n\n"
python3 $METAXCAN/PrediXcanAssociation.py \
--expression_file $RESULTS/predixcan/Whole_Blood__predict.txt \
--input_phenos_file $DATA/phenotype/random_pheno_1000G_hg38.txt \
--input_phenos_column pheno \
--output $RESULTS/predixcan/random_pheno/Whole_Blood__association.txt \
--verbosity 9 \
--throw
```
## read results
```{r, eval=FALSE}
predixcan_association = read_tsv(glue::glue("{results.dir}/predixcan/random_pheno/Whole_Blood__association.txt"))
dim(predixcan_association)
predixcan_association %>% arrange(pvalue) %>% head
predixcan_association %>% arrange(pvalue) %>% ggplot(aes(pvalue)) + geom_histogram(bins=20)
```
## Exercise
-[ ] Run association with another phenotype
in $PRE/predixcan/data/phenotype/ALL.chr22.shapeit2_integrated_snvindels_v2a_27022019.GRCh38_x_en_Whole_Blood.simulated_phenotype.spike_n_slab_0.5_x_pve0.6.txt
```{bash, eval=FALSE}
```
```{r}
```
-------
-------
# Summary PrediXcan
```{r}
## harmonized and imputed GWAS result for coronary artery disease is available in
# $PRE/s-predixcan/data/
```
## run s-predixcan
```{bash, eval=FALSE}
export PRE="/home/rstudio/QGT-Columbia-HKI"
export DATA=$PRE/s-predixcan/data
export MODEL=$PRE/models
export METAXCAN=$PRE/repos/MetaXcan-master/software
export RESULTS=$PRE/results
python $METAXCAN/SPrediXcan.py \
--gwas_file $DATA/imputed_CARDIoGRAM_C4D_CAD_ADDITIVE.txt.gz \
--snp_column panel_variant_id --effect_allele_column effect_allele --non_effect_allele_column non_effect_allele --zscore_column zscore \
--model_db_path $MODEL/gtex_v8_mashr/mashr_Whole_Blood.db \
--covariance $MODEL/gtex_v8_mashr/mashr_Whole_Blood.txt.gz \
--keep_non_rsid --additional_output --model_db_snp_key varID \
--throw \
--output_file $RESULTS/spredixcan/eqtl/CARDIoGRAM_C4D_CAD_ADDITIVE__PM__Whole_Blood.csv
```
## plot and interpret s-predixcan results
```{r, eval=FALSE}
spredixcan_association = read_csv(glue::glue("{results.dir}/spredixcan/eqtl/CARDIoGRAM_C4D_CAD_ADDITIVE__PM__Whole_Blood.csv"))
dim(spredixcan_association)
spredixcan_association %>% arrange(pvalue) %>% head
spredixcan_association %>% arrange(pvalue) %>% ggplot(aes(pvalue)) + geom_histogram(bins=20)
```
SORT1, considered to be a causal gene for LDL cholesterol and as a consequence of coronary artery disease, is not found here. Why? (tissue)
## run multixcan (optional)
```{bash, eval=FALSE}
python $METAXCAN/SMulTiXcan.py \
--models_folder $DATA/models/eqtl/mashr \
--models_name_pattern "mashr_(.*).db" \
--snp_covariance $DATA/models/gtex_v8_expression_mashr_snp_covariance.txt.gz \
--metaxcan_folder $RESULTS/spredixcan/eqtl/ \
--metaxcan_filter "CARDIoGRAM_C4D_CAD_ADDITIVE__PM__(.*).csv" \
--metaxcan_file_name_parse_pattern "(.*)__PM__(.*).csv" \
--gwas_file $RESULTS/processed_summary_imputation/imputed_CARDIoGRAM_C4D_CAD_ADDITIVE.txt.gz \
--snp_column panel_variant_id --effect_allele_column effect_allele --non_effect_allele_column non_effect_allele --zscore_column zscore --keep_non_rsid --model_db_snp_key varID \
--cutoff_condition_number 30 \
--verbosity 7 \
--throw \
--output $RESULTS/smultixcan/eqtl/CARDIoGRAM_C4D_CAD_ADDITIVE_smultixcan.txt
```
# Colocalization methods
## fine-map GWAS results -
We will run torus due to time limitation but ideally we would like to run a method that allows multiple causal variants per locus.
```{bash, eval=FALSE}
#torus -d Height.torus.zval.gz --load_zval -dump_pip Height.gwas.pip
#gzip Height.gwas.pip
torus -d /Users/haekyungim/Box/LargeFiles/QGT-Columbia-HKI/fastenloc/data/Height.torus.zval.gz --load_zval -dump_pip /Users/haekyungim/Box/LargeFiles/QGT-Columbia-HKI/fastenloc/data/Height.gwas.pip
gzip Height.gwas.pip
```
## estimate priors
is this done internally by fastENLOC?
```{r}
```
## calculate colocalization with fastENLOC
```{bash, eval=FALSE}
## tutorial https://github.com/xqwen/fastenloc/tree/master/tutorial
export EQTLGZ=eqtl_annotation_gzipped
export GWASGZ=gwas_data_gzipped
export TISSUE=Whole_Blood
fastenloc -eqtl EQTLGZ -gwas GWASGZ -t tissue_name #[-total_variants total_snp] [-thread n] [-prefix prefix_name] [-s shrinkage]
```
## analyze results
```{r}
## optional - compare with s-predixcan results
```
# Mendelian randomization methods
## run SMR (optional)
```{bash, eval=FALSE}
```
## run TWMR (for a locus)
```{bash, eval=FALSE}
cd $TWMR
export PRE=/Users/haekyungim/Box/LargeFiles/QGT-Columbia-HKI
export TWMR=$PRE/repos/TWMR-master
export OUTPUT=$PRE/results
GENE=ENSG00000002919
R < $TWMR/MR.R --no-save $GENE
cd $PRE
```