Correction: Integrative genomic analysis of the human immune response to influenza vaccination

Franco LM, Bucasas KL, Wells JM, Niño D, Wang X, Zapata GE, Arden N, Renwick A, Yu P, Quarles JM, Bray MS, Couch RB, Belmont JW, Shaw CA. 2013. Integrative genomic analysis of the human immune response to influenza vaccination. eLife 2:e00299. doi: 10.7554/eLife.00299.

Published July 16, 2015

In our manuscript concerning integrative genomic analysis of the response to influenza vaccination in humans, we included a section concerning perturbation eQTLs. Perturbation eQTLs are associations of gene expression with genotype that may change in magnitude, direction, or only appear after exposure of individuals to experimentally modifiable covariates or variable environmental conditions. In the case of our study, influenza vaccination serves as a perturbation exposure, and gene expression profiling was performed serially on subjects before and after the vaccination. This design permits the investigation of perturbation eQTLs.

Our paper presented an expository of a perturbation eQTL using the gene NECAB2 (Figure 2A), as well as more global analysis of perturbation eQTLs characteristics in our data (Figure 2B). In error, we presented two identical Manhattan plot graphics for NECAB2 in Figure 2A, both from male subjects instead of graphics for the distinct male and female cohorts in our study. Upon review, the male plot strongly shows the induction of an association signal after exposure, but the female cohort shows an association signal at baseline. Therefore, NECAB2 does not replicate the pattern that would be expected of a perturbation eQTL. Nevertheless, perturbation eQTLs are potential contributors to association patterns in our study. This was illustrated in our global analysis, which showed an increase in the magnitude of the genotype effect (R2g), explained by an increase in beta without evidence of decrease in the mean within genotype variance (presented in Figure 2B). Our objective in presenting NECAB2 was to provide a motivating example, and NECAB2 as a perturbation was not otherwise important to the conclusions in our work. The correct pair of Manhattan plots for NECAB2 has been included here (Correction figure 1).

Correction figure 1

Download asset Open asset

To clarify the potential contribution of perturbation eQTLs, we present additional expository examples of the genes ADCY3 (Correction figure 2) and DISC1 (Correction figure 3). These genes present a replicated perturbation eQTL pattern in both the male and female data.

Correction figure 2

Download asset Open asset

Correction figure 3

Download asset Open asset

We appreciate the opportunity to correct the error made in presenting the graphics of NECAB2. The corrected version of Figure 2 is shown below:

The originally published Figure 2 is also shown for reference:

More research in this area is warranted, and we hope other studies will examine the potential of this distinctive form of gene–environment interaction.

The article has been corrected accordingly.