Bidirectional Trait Mendelian Randomization
Review phenotype-oriented MR evidence with the same search-to-results flow as the eQTL workspace.
Dataset Catalog
Available bidirectional MR result sets
Select a cohort. GWAS datasets map to FinnGen traits; FinnGen datasets map to UKB Neale traits.
Result Review
Review the bidirectional MR table and figure set.
MR Table
Current filtered bidirectional records
Direction depends on available genetic instruments. If a neuroendocrine neoplasms endpoint has no SNPs passing the selected instrument threshold, it is retained only as the outcome and exposure-to-NENs records are shown.
Sensitivity analyses follow the number of instruments: records with one SNP are not tested, two SNPs report heterogeneity only, three SNPs report heterogeneity and pleiotropy, and four or more SNPs report the full set. Although MR-PRESSO can be computed at nsnp = 3, leave-one-out leaves only two IVs, so outlier detection and the Distortion test are not statistically meaningful; this study reports the MR-PRESSO global test P value only when nsnp >= 4.
| Direction | Feature | Method | Beta | SE | or | or_lci95 | or_uci95 | pval | fdr | heterogeneity_pval | pleiotropy_pval | mrpresso_pval | nsnp | NENs Cancer Type |
|---|
Figure Panel
Interpretation figures for the active bidirectional MR result set
Double-click any figure to enlarge it. Each card provides both PNG and PDF downloads. The forest plot and heatmap share the same default 12-feature view, while the volcano plot can accent selected features from the current result set.
In the sensitivity heatmap, values below 0.05 are shown using a distinct highlight color, whereas values from 0.05 to 1 are mapped to the shared red-to-blue legend scale.
Example: AB1_ACTINOMYCOSIS, SCLC, COLON_POLYP
Feature label color: NENs as exposure vs NENs as outcome
Default sync: forest selection drives this panel
AI Overview
AI Result Interpretation
This panel uses the configured AI service to analyze results and answer questions. If AI chat is disabled, use Export AI report to obtain the standardized output and submit it to an LLM.
Click Send to interpret the current analysis result.