Fig. 4

Distribution of a observed count values for titration-specific (TS) features and b expected count values for unmixed-specific (US) features by pipeline and individual. The orange horizontal dashed line indicates a count value of 1. c Artifactual feature proportion (Art. Feat. Prop.) for titration-specific and dunmixed-specific features with an adjusted pvalue < 0.05 for the Bayesian hypothesis test and binomial test respectively. We failed to accept the null hypothesis when the p value < 0.05, indicating that the discrepancy between the feature only being observed in the titrations or unmixed samples cannot be explained by sampling alone