Extended Data Fig. 7: Independent validation of genes with limited prior evidence.
From: Large-scale exome analyses reveal new rare variant contributions in amyotrophic lateral sclerosis
Forest plots comparing odds ratios (OR; center) with 95% confidence intervals (CI; error bars) for main analyses versus sensitivity analyses that exclude potential overlap with prior studies. a, Plot comparing the ARPP21 p.P563L association from the full dataset to an analysis excluding four carriers who could potentially overlap with previous family studies. b, Plot comparing the CFAP410 p.V58L association in the full discovery (ncases = 11,763; ncontrols = 69,677) and replication (ncases = 4,781; ncontrols = 130,927) cohorts to a sensitivity analysis that excludes duplicated or genetically related individuals from the discovery (ncases = 4,474; ncontrols = 65,643) and replication (ncases = 3,698; ncontrol = 130,802) sets. c, Plot comparing the DNAJC7 URV association in the full discovery (ncases = 13,138; ncontrols = 69,775) and replication (ncases = 4,781; ncontrols = 130,928) cohorts to a sensitivity analysis that excludes individuals from cohorts that potentially overlap with the original study from the discovery (ncases = 7,606; ncontrols = 59,926) and replication (ncases = 4,591; ncontrols = 130,928) sets. Association statistics were estimated using Firth’s logistic regression with profile penalized likelihood confidence intervals.
