Depression_GI

This code relates to the project titled "Depression and risk of gastrointestinal disorders: a comprehensive two-sample Mendelian randomization study".

Step 1: Calculate R2 and F-statistic

First, we needed to calculate R2 and F-statistic for MDD instrumental variables. We downloaded complete summary statistics for MDD from PGC.

1_calculate_R2&Fstatistic.R

Step 2: Find LD proxies for missing SNPs

Next, we needed to find linkage disequilibrium (LD) proxies for genetic variants missing in the outcome dataset. Here, we did this using the LDlinkR package.

2_find_LD_proxies_for_MainConsortium.R

Step 3: Harmonisation

This step conduct Harmonisation for UK Biobank and FinnGen outcome, produce instrumental variables Output in two-sample MR format.

3.1_harmonisation_for_MainConsortium.R
3.2_harmonisation_for_FinnGen.R

Step 4: Run two-sample MR analysis

Once our exposure and outcome datasets were ready, we ran the two-sample MR analysis. We did this separately for UK Biobank and FinnGen.

4_run_2SMR_for_MainConsortium.R
4_run_2SMR_for_FinnGen.R

Step 5: Run meta-analysis

We combined two-sample MR results obtained in the previous step using a fixed-effect meta-analysis. This was done using the meta package.

5_Mete_analysis.R

Step 6: Sensitity Analyses

This step focuses on performing sensitivity analysis, including "Remove_pleiotrpy_snps", "CAUSE", "risk_factor_outcome" and "negative_control_outcome".

6_SensitityAnalyses_negative_control_outcome.R
6_SensitityAnalyses_risk_factor_outcome.R
6_SensitityAnalyses_run_2SMR_for_FinnGen_after_remove_pleiotrpy_snps.R
6_SensitityAnalyses_run_2SMR_for_MainConsortium_after_remove_pleiotrpy_snps.R
6_SensitityAnalyses_run_CAUSE_for_MDD_GERD.R
6_SensitityAnalyses_run_CAUSE_for_MDD_IBS.R
6_SensitityAnalyses_run_CAUSE_for_MDD_NAFLD.R
6_SensitityAnalyses_run_CAUSE_for_MDD_PUD.R

Step 7: Secondary Analyses

This step focuses on performing sensitivity analysis, including "Reverse MR", "LDSC" and "MVMR".

7_SecondaryAnalyses_LDSC.sh
7_SecondaryAnalyses_MVMR.R
7_SecondaryAnalyses_Reversed_direction_find_LD_proxies.R
7_SecondaryAnalyses_Reversed_direction_harmonisation.R
7_SecondaryAnalyses_Reversed_direction_run_2SMR.R

Step 8: Create visualisations

This step create graphics diagnostics visualisation for main outcome two-sample MR including scatter plot, funnel plot, forest plot and leave-one-out plot. We created forest plots for discovery, replication and meta analyses using the forestplot packages.

8_Create_graphics_diagnostics_visualisations.R
8_ForestPlot.R