/MR_tricks

Primary LanguageR

Post-GWAS analysis and Mendelian Randomization

During the study based on Mendelian Randomization, several functions were proposed to solve several common problems, including:

Special note

Recently, I noticed that some uploaders in Bilibili seem to sell my codes, such as Shining94. The original purpose of this project is to share my solutions to certain problems in Mendelian Randomization without charge. I urge those guys stop selling my codes, and I decide to not share my recent findings on MR. If you have any questions, leave me a message.

Preprocessing for exposure GWASs

  • how to get rsID from chromosome and position: get_rsID_from_chr_pos
  • how to find allele frequency, proxy of a given SNP: snp_add_eaf, snp_add_eaf_manual,find_proxy
  • how to search pleiotropy for a given SNP: snp_phenoscanner

Exposure and outcome data harmonization

  • how to harmonize if direct removal is not allowed and proxy is required: snp_replace_proxy, harmonise_data_modified

Main analysis for Mendelian Randomization

  • mr_meta is helpful conduct meta-analysis to pool estimates from different cohorts
  • mr_modified is alternative function to mr in TwoSampleMR package, since the mr-raps cannot not be conducted in the original function
  • get_cohort_from_number transforms a number into a name of cohort
  • get_method_from_abbr transforms an abbreviation version of MR method into a full name
  • get_protein_from_uniprot transforms a UniProt ID to full name of protein, with UniProtRefPanel.xlsx as the reference panel

More functions are under developing...

  • data visualization: volcano plot, forest plot...
  • colocolization
  • TWAS
  • ...

Useful references