Regression with Summary Statistics (RSS) Fine-Mapping and TWAS with SuSiE

Regression with Summary Statistics (RSS) Fine-Mapping and TWAS with SuSiE#

Description#

Last, we include an option to conduct fine-mapping with SuSiE Regression using Summary Statistics (RSS) model and TWAS.

Input#

--ld-meta-data: file with chrom, start, end and path columns. For example:

#chrom  start   end     path
chr1    101384274       104443097       chr1/chr1_101384274_104443097.cor.xz,chr1/chr1_101384274_104443097.cor.xz.bim
chr1    104443097       106225286       chr1/chr1_104443097_106225286.cor.xz,chr1/chr1_104443097_106225286.cor.xz.bim
chr1    106225286       109761915       chr1/chr1_106225286_109761915.cor.xz,chr1/chr1_106225286_109761915.cor.xz.bim
chr1    109761915       111483530       chr1/chr1_109761915_111483530.cor.xz,chr1/chr1_109761915_111483530.cor.xz.bim

--gwas-meta-data: file with information on GWAS. For example:

study_id        chrom   file_path       column_mapping_file     n_sample        n_case  n_control
AD_Bellenguez_2022      0       /data/GWAS/AD_GWAS/GCST90027158_buildGRCh38.tsv.gz      /data/GWAS/column_mapping_file/Bellenguez.yml   0       111326  677663
AD_Jansen_2021  0       /data/GWAS/AD_GWAS/AD_sumstats_Jansenetal_2019sept.hg38.txt     /data/GWAS/column_mapping_file/Jansen.yml       0       71880   383378
AD_Kunkle_Stage1_2019   0       /data/GWAS/AD_GWAS//Kunkle_etal_Stage1_results.txt_file_1_hg38.txt      /data/GWAS/column_mapping_file/Kunkle_stage_1.yml       0       21982   41944
AD_Wightman_Full_2021   0       /data/GWAS/AD_GWAS/PGCALZ2full.hg38.txt /data/GWAS/column_mapping_file/AD_Wightman_Full_2021.yml0       90338   1036225

--qc_method: set to rss_qc, dentist, or slalom.

--finemapping_method: set to single_effect, susie_rss, or bayesian_conditional_regression.

--cwd: output path

--skip_analysis_pip_cutoff: defaults to 0.025

--skip_regions: format as chr:start-end. For example: 6:25000000-35000000

--region_name: format as chr:start-end. For example: 22:49355984-50799822

Minimal Working Example Steps#

v. Run the Summary Statistics Fine-Mapping#

sos run $PATH/rss_analysis.ipynb univariate_rss \
--ld-meta-data $PATH/ldref/ld_meta_file.tsv \
    --gwas-meta-data $PATH/GWAS_sumstat_meta_cloud_Apr_9.tsv \
    --qc_method "rss_qc" --impute \
    --finemapping_method "susie_rss" \
    --cwd $PATH/output/ \
    --skip_analysis_pip_cutoff 0 \
    --skip_regions 6:25000000-35000000 \
    --region_name 22:49355984-50799822

Anticipated Results#

Summary statistics fine-mapping produces a results file for each region and gwas of interest.

RSS_QC_RAISS_imputed.chr22_49355984_50799822.univariate_susie_rss.rds:

  • For each region in region_name and gwas in the gwas-meta-data file:

  1. RSS_QC_RAISS_imputed:
    a. variant_names
    b. analysis_script
    c. sumstats
    d. susie_result_trimmed
    e. outlier_number

a. chrom
b. pos
c. variant_id
d. A1
e. A2
f. var
g. raiss_ld_score
h. raiss_R2
i. pvalue
j. effect_allele_frequency
k. odds_ratio
l. ci_lower
m. ci_upper
n. beta
o. se
p. n_case
q. n_control
r. het_isq
s. het_pvalue
t. variant_alternate_id
u. z

A file for each gwas in gwas-meta-data like: RSS_QC_RAISS_imputed.chr22_49355984_50799822.AD_Bellenguez_EADB_2022.sumstats.tsv.gz. The contents of these are included in the .rds file above

Contents