Correction: Replication Study: Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia

Shan X, Fung JJ, Kosaka A, Danet-Desnoyers G, Reproducibility Project: Cancer Biology. 2017. Replication Study: Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. eLife 6:e25306. doi: 10.7554/eLife.25306.

Published 27, June 2017

This is a corrigendum for Replication Study: Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia, which includes the following corrections:

First, the direction of the effect size calculated for the Replication Study’s comparison of fold BCL2 expression in K-562 cells compared to a DMSO constant of 1 (Figure 5A, top panel) is incorrect. The direction of the effect size (Cohen’s d) for this comparison should be d = −2.04 95% CI [−4.17, 0.13] (instead of the reported d = 2.04 95% CI [−0.13, 4.17]), which results in a change in the meta-analysis. The meta-analysis effect size should be d = −0.307, 95% CI [−3.57, 2.96], p =0.854 (instead of the reported d = 1.58, 95% CI [0.36, 2.79], p=0.011).

Second, the calculation of variance in the meta-analysis of mice treated daily with I-BET151 compared to vehicle control mice (Figure 5C) is incorrect. Subsequently, the effect size and the associated p value of the meta-analysis are incorrect. The effect size (Hazard Ratio) for this comparison should be HR = 3.30 95% CI [0.49, 22.18] (instead of the reported HR = 6.13 95% CI [1.11, 34.02]) resulting in a p value of p=0.220 (instead of the reported p = 0.038).

Finally, the language surrounding these comparisons in the Results and Discussion section as well as in Figure 5 and the Figure Legend (Figure 5) have been amended to account for these corrections. Additional r code for the meta-analysis can be found on the Open Science Framework: https://osf.io/vfp47/.

The corrected Figure 5 is shown here:

The originally published Figure 5 is also shown for reference:

The article has now been corrected.