Nature 480, 118–122 (2011); doi:10.1038/nature10598

In the Methods section of this Letter, the description of immunohistochemical (IHC) analysis did not clearly describe the scoring system we used. To avoid confusion, the assignment of proportion scores should have been described as follows: “0 if 0% of the tumour cells showed positive staining, 0.1–1.0 if 0.1% to 1% of cells were stained, 1.1–2.0 if 1.1% to 10% stained, 2.1–3.0 if 11% to 30% stained, 3.1–4.0 if 31% to 70% stained, and 4.1–5.0 if 71% to 100% stained. (Each percentage range of the stained tumour section is further divided into ten smaller percentage ranges and represented by corresponding non-integral scores, such as 3.1 representing 31%–34%). We rated the intensity of staining on a scale of 0 to 3: 0, negative; 1, weak; 2, moderate; and 3, strong. We then combined the proportion and intensity scores to obtain a total score (range, 0–8), as described previously.”

In addition, in Supplementary Fig. 20, the x and y axes labels of each of the three graphs were inadvertently reversed. A revised figure is included as Supplementary Information to this Corrigendum. The data in all panels remain unchanged, and these errors do not affect the results or conclusions of the Letter.

We apologize any confusion this may have caused. The original Letter has not been corrected.