Suggestions on how to select genetic variants to use as instrumental variables

• Identify genetic variants from candidate gene studies and genome-wide association studies (GWAS):

1)      P thresholds for association of both 10^-8 have been suggested.

2)      Information on monogenic disorders: https://www.omim.org/.

3)      Summary level findings from GWAS: http://www.gwascentral.org/

• Check if selected genetic variants are in linkage disequilibrium.

1)      SNiPA: http://snipa.helmholtz-muenchen.de/snipa3/

2)      LD Hub: http://ldsc.broadinstitute.org/

3)      Genetic variants in major linkage disequilibrium should be excluded.

• Check whether selected genetic variants have pleiotropic effects.

1)      LD Hub: http://ldsc.broadinstitute.org/

2)      Pheno Scanner: http://www.phenoscanner.medschl.cam.ac.uk/phenoscanner

3)      GWAS-Central: http://www.gwascentral.org/can be searched by rs-number

4)      Genetic variants can be annotated to genes using:

5)      SNiPA: http://snipa.helmholtz-muenchen.de/snipa3/

6)      SNPedia: https://www.snpedia.com/index.php/SNPedia

7)      Ensembl: http://www.ensembl.org/index.html

8)      UCSC Genome Browser: https://genome.ucsc.edu/cgi-bin/hgGateway

• Check genetic variants for deviation from the Hardy-Weinberg expectations as an indication of a genotyping error or a violation of Mendel’s second law. Check that allele frequencies largely correspond to frequencies reported in genomic databases, i.e.:

1)      http://www.ensembl.org/index.html

2)      https://genome.ucsc.edu/cgi-bin/hgGateway

3)      https://www.snpedia.com/index.php/SNPedia

• Link to tool for two-sample summary level Mendelian randomization

1)      http://www.mrbase.org/