Fig. 1From: Mendelian randomization analyses of known and suspected risk factors and biomarkers for myasthenia gravis overall and by subtypesThe primary design and assumptions of the bidirectional Mendelian randomization research: (1) instrumental variables are closely linked with exposure, (2) instrumental variables and any confounding factors were independent, and (3) instrumental variables could influence outcomes only via exposure. SNP: single-nucleotide polymorphismBack to article page