Sovereign 8s.is mirror of standardgalactic/TS2024_UKBBageEffect
- R 100%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
| analysis | ||
| figures | ||
| README.md | ||
Design and model choices shape inference of age-varying genetic effects on complex traits
1. Preprocessing of phenotype data
- Analytical script: extractPheno.R
- All necessary functions required to run the script are included in functions.R
- Extract and recode phenotype data
- Generate slopes of change for genome-wide tests on change
2. Perform genome-wide scans using REGENIE
GWA in REGENIE - Step 1:
$myprog \
--step 1 \
--bed $DATA/ukbb/_001_ukb_cal_allChrs \
--phenoFile $OUT/output/regenie/phenofiles/$run \
--keep $DATA/ukbb/qc_pass_MAC100.id \
--extract $DATA/ukbb/qc_pass_MAC100.snplist \
--force-qt \
--minINFO 0.9 \
--phenoCol $run \
--covarFile $OUT/output/regenie/phenofiles/${run}_Covar \
--catCovarList SEX,batch \
--maxCatLevels 3000 \
--bsize 1000 \
--lowmem \
--lowmem-prefix $OUT/output/regenie/genofiles/${run} \
--out $OUT/output/regenie/genofiles/${run}_s1 \
--threads $SLURM_CPUS_PER_TASK
GWA in REGENIE - Step 2:
$myprog \
--step 2 \
--pgen $UKBB/genotypes/imp/v3/pgen/ukb22828_c"$a"_b0_v3 \
--ref-first \
--phenoFile $OUT/output/regenie/phenofiles/$run \
--covarFile $OUT/output/regenie/phenofiles/${run}_Covar \
--catCovarList SEX,batch \
--maxCatLevels 3000 \
--phenoCol $run \
--approx \
--pred $OUT/output/regenie/genofiles/${run}_s1_pred.list \
--bsize 400 \
--out $OUT/output/regenie/genofiles/chr${a}_${run} \
--force-qt \
--threads $SLURM_CPUS_PER_TASK \
$extra_args
Parameter specifications
- gene-by-age-interaction model (cross-sectional):
extra_args="--interaction age --no-condtl" - phenotype outcome:
$run(e.g, run=bmi_0 for baseline bmi, run=bmi_change for change in bmi) - REGENIE software:
myprog=$PROGS/regenie/regenie_v3.2.6.gz_x86_64_Linux - chromosome number:
$a
Process the results
-
Analytical script: processGWA.R
- generates summary statistic files
- performs clumping to select LD-independent variants
-
Analytical script: sumGWA.R
- summarizes clumping output for all traits and models
3. Processing of variant-specific data
-
Analytical script: gene.R
- Extract defined genetic variants for UKB participants
- Required R-packages: bigsnpr
- Estimate genetic effects across age-heterogeneous subgroups per variant
4. Generate participation weights
- Analytical script: bias.R
- Performs generate Inverse Probability Weights to correct for selective participation in UKB follow-ups
5. Generate polygenic scores
- Analytical script: pgs.R
- Generates polygenic scores to test for polygenic x age interaction effects step 6
- Requires the polygenic score software PRSice-2
6. Obtain age estimates and age-varying genetic effects for candidate genetic variants
-
Analytical script: lme.R
- Use for phenotypic analyses, including cross-sectional and longitudinal age effects, cohort effects, quadratic age effects and weighted age effects
- Use for genotypic analyses, including cross-sectional and longitudinal age-varying genetic effects, cohort-varying genetic effects, quadratic age-varying genetic effects and weighted age-varying genetic effects
7. Perform Mendelian Randomization analysis using time-varying genetic effects
- Analytical script: mr.R
-
- Required R-packages: TwoSampleMR
8. Summary of the results
-
Analytical script: summaryLongitudinal.R
- Script to summarize the results and generate manuscript plots
- All necessary functions required to run the script are included in summaryAgeFunc.R
- All supplement tables can be generated using createTablesAge.R