Mendelian randomization using semiparametric linear transformation models

Mediation and instrumental variable analyses for vaccine-induced antibody titer against influenza B

OBJECTIVE

Immune correlate analyses for vaccine trials have been applied to investigate associations of vaccine efficacy and surrogate markers such as vaccine-induced antibodies. However, the role of antibody as a surrogate marker in predicting the outcome can vary by time, and surrogate-outcome confounding may have resulted in bias even in randomized trials. We provide a framework for surrogate marker assessment to address the aforementioned issues.

STUDY DESIGN AND SETTING

We reanalyzed the vaccine randomized trial for influenza B. We conducted a mediation analysis that enables estimation of vaccine efficacy, mediation effects and proportion of mediation on disease probabilities at various follow-up times. We proposed instrumental variable (IV) analyses with randomized vaccination as an IV accounting for potential unmeasured confounding.

RESULTS

The mediation effect of vaccine efficacy by hemagglutination inhibition (HAI) titer was significantly protective at 181 days after vaccination: 63.2% [95% confidence interval, (CI) = (39.9%, 82.0%)], and HAI titer explained 61.1% [95% CI = (36.7%, 96.2%)] of the protective effect of vaccination.

CONCLUSIONS

Most of vaccine efficacy is mediated by HAI titer, particularly in children 10 years and older. Our contribution is to provide causal analytics for the role of surrogate marker with weaker assumptions regarding surrogate-disease causation.

Efficient gradient boosting for prognostic biomarker discovery

MOTIVATION

A gradient boosting decision tree (GBDT) is a powerful ensemble machine-learning method that has the potential to accelerate biomarker discovery from high-dimensional molecular data. Recent algorithmic advances, such as extreme gradient boosting (XGB) and light gradient boosting (LGB), have rendered the GBDT training more efficient, scalable and accurate. However, these modern techniques have not yet been widely adopted in discovering biomarkers for censored survival outcomes, which are key clinical outcomes or endpoints in cancer studies.

RESULTS

In this paper, we present a new R package 'Xsurv' as an integrated solution that applies two modern GBDT training frameworks namely, XGB and LGB, for the modeling of right-censored survival outcomes. Based on our simulations, we benchmark the new approaches against traditional methods including the stepwise Cox regression model and the original gradient boosting function implemented in the package 'gbm'. We also demonstrate the application of Xsurv in analyzing a melanoma methylation dataset. Together, these results suggest that Xsurv is a useful and computationally viable tool for screening a large number of prognostic candidate biomarkers, which may facilitate future translational and clinical research.

AVAILABILITY AND IMPLEMENTATION

'Xsurv' is freely available as an R package at: https://github.com/topycyao/Xsurv.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.