Abstract
Following the publication of our recent article (Kapp et al., BMC Genomics 2006, 7:231), we (the authors) regrettably found several errors in the published Table 5. This correction article not only describes what makes the published Table 5 incorrect, it also presents the correct Table 5.
I regret to say that several errors exist in Table 5 of our recently published paper [1]. Here we present the correct version of Table 5 and explain what is incorrect about the published Table 5. (See Table 1 of this manuscript.)
Table 1.
The (original) incorrect Table 5 as published in Kapp et al. (2006).
| BCMP11/ABCC11 | ||||
| Group 1 | Group 2 | Group 3 | ||
| SLC39A6/GATA3 | Group 1 | 42 | 0 | 6 |
| Group 2 | 8 | 21 | 3 | |
| Group 3 | 0 | 0 | 16 | |
| Sørlie subtype | Normal-like | 6 | 0 | 0 |
| ERBB2+ | 4 | 7 | 0 | |
| Luminal A | 28 | 0 | 0 | |
| Luminal B | 9 | 2 | 0 | |
| Basal | 5 | 8 | 6 | |
A typographical error exists in Table 5 (upper panel). The 6 that is presented in the upper right-hand corner of the published Table 5 (upper panel) should be a zero. The correct version of Table 5 (upper panel) is shown in Table 2 of the present manuscript.
Table 2.
The correct Table 5 (upper panel).
| BCMP11/ABCC11 | ||||
| Group 1 | Group 2 | Group 3 | ||
| SLC39A6/GATA3 | Group 1 | 42 | 0 | 0 |
| Group 2 | 8 | 21 | 3 | |
| Group 3 | 0 | 0 | 16 | |
The paragraphs in the paper associated with Table 5 (upper panel) are correct as published.
Table 5 (lower panel) has a more significant error. The correct version of Table 5 (lower panel) is shown in Table 3 of the present manuscript.
Table 3.
The correct Table 5 (lower panel).
| BCMP11/ABCC11 | ||||
| Group 1 | Group 2 | Group 3 | ||
| Sørlie subtype | ERBB2+ | 1 | 6 | 0 |
| Luminal A | 15 | 0 | 0 | |
| Luminal B | 3 | 0 | 0 | |
| Basal | 0 | 1 | 8 | |
| None | 11 | 0 | 0 | |
What was presented in the paper was a comparison of the Sørlie et al. (2003) classifications of the Sørlie dataset (without a cutoff), not a subset of the Norway/Stanford arrays.
Moreover, the Comparison of ESR1/ERBB2 subtypes and Sørlie et al. (2003) subtypes section is incorrect. The text in this section is associated with Table 5 (lower panel) and describes the results for the Sørlie dataset, not the results for a subset of the Norway/Stanford dataset as intended (and as stated in the table's legend). Based upon the correct Table 5 (lower panel) which presents the results for the Norway/Stanford arrays of samples from Ullevål University Hospital, the subtypes described by [2] are fairly similar to the ESR1/ERBB2 subtypes we defined. All of the luminal A and luminal B samples were classified to Group 1 (the ESR1+/ERBB2- subtype). All but one of the Ullevål samples in the Norway/Stanford dataset that were classified to the ERBB2-overexpressing Sørlie et al. (2003) subtype were classified by us to Group 2 (the ERBB2+ subtype). Finally, except for one sample, all of the basal samples were classified to Group 3 (the ESR1-/ERBB2- subtype). All of the samples that were not classified to any of the Sørlie et al. (2003) subtypes were classified to our ESR1+/ERBB2- subtype rather than being uniformly distributed among all three of our subtypes.
It may be surprising to see one of the basal samples in our ERBB2+subtype and not with the rest of the basal samples in our ESR1-/ERBB2- subtype. In all other papers, the basal subtype has been the most cohesive. Our ESR1/ERBB2 centroids consist of approximately four times as many genes as the Sørlie et al. (2003) centroids do (1908 genes and 496 genes, respectively). Not all of the genes in the ESR1/ERBB2 centroids belong to the ESR1 gene cluster or to the ERBB2 gene cluster. It is the influence of the genes which do not belong to these clusters that causes one of the basal samples to not be classified with the others.
Table 4 of the present document contains the 45 arrays used to make the correct Table 5 (lower panel). Anita Langerød classified each array to one of the Sørlie et al. (2003) subtypes if the array's correlation with one of the subtypes centroids was at least 0.2. These classifications will appear in an upcoming publication.
Table 4.
The arrays used to make the correct Table 5 (lower panel)
| Array | Sørlie et al. (2003) subtype | ESR1/ERBB2 subtype |
| ULL-D-002 | Luminal A | ESR1+/ERBB2- |
| ULL-D-016 | Luminal A | ESR1+/ERBB2- |
| ULL-D-020 | Luminal A | ESR1+/ERBB2- |
| ULL-D-022 | Luminal A | ESR1+/ERBB2- |
| ULL-D-023 | Luminal B | ESR1+/ERBB2- |
| ULL-D-037 | Luminal A | ESR1+/ERBB2- |
| ULL-D-044 | Luminal A | ESR1+/ERBB2- |
| ULL-D-048 | Luminal A | ESR1+/ERBB2- |
| ULL-D-056 | Luminal B | ESR1+/ERBB2- |
| ULL-D-057 | Basal-like | ESR1-/ERBB2- |
| ULL-D-066 | None | ESR1+/ERBB2- |
| ULL-D-071 | ERBB2 | ERBB2+ |
| ULL-D-075 | Basal-like | ERBB2+ |
| ULL-D-080 | Basal-like | ESR1-/ERBB2- |
| ULL-D-083 | ERBB2 | ERBB2+ |
| ULL-D-085 | None | ESR1+/ERBB2- |
| ULL-D-113 | Luminal B | ESR1+/ERBB2- |
| ULL-D-134 | Luminal A | ESR1+/ERBB2- |
| ULL-D-150 | Luminal A | ESR1+/ERBB2- |
| ULL-D-165 | None | ESR1+/ERBB2- |
| ULL-D-167 | Basal-like | ESR1-/ERBB2- |
| ULL-D-169 | ERBB2 | ERBB2+ |
| ULL-D-177 | Basal-like | ESR1-/ERBB2- |
| ULL-D-184 | None | ESR1+/ERBB2- |
| ULL-D-007 | None | ESR1+/ERBB2- |
| ULL-D-011 | Luminal A | ESR1+/ERBB2- |
| ULL-D-013 | Luminal A | ESR1+/ERBB2- |
| ULL-D-026 | Basal-like | ESR1-/ERBB2- |
| ULL-D-027 | ERBB2 | ERBB2+ |
| ULL-D-038 | None | ESR1+/ERBB2- |
| ULL-D-053 | None | ESR1+/ERBB2- |
| ULL-D-065 | Basal-like | ESR1-/ERBB2- |
| ULL-D-067 | None | ESR1+/ERBB2- |
| ULL-D-074 | Luminal A | ESR1+/ERBB2- |
| ULL-D-087 | ERBB2 | ESR1+/ERBB2- |
| ULL-D-096 | ERBB2 | ERBB2+ |
| ULL-D-099 | Basal-like | ESR1-/ERBB2- |
| ULL-D-101 | ERBB2 | ERBB2+ |
| ULL-D-122 | Luminal A | ESR1+/ERBB2- |
| ULL-D-132 | None | ESR1+/ERBB2- |
| ULL-D-135 | None | ESR1+/ERBB2- |
| ULL-D-139 | Basal-like | ESR1-/ERBB2- |
| ULL-D-143 | Luminal A | ESR1+/ERBB2- |
| ULL-D-144 | Luminal A | ESR1+/ERBB2- |
| ULL-D-183 | None | ESR1+/ERBB2- |
Contributor Information
Amy V Kapp, Email: AKapp@stanford.edu.
Stefanie S Jeffrey, Email: ssj@stanford.edu.
Anita Langerød, Email: anitala@ulrik.uio.no.
Anne-Lise Børresen-Dale, Email: alb@radium.uio.no.
Wonshik Han, Email: HanW@snu.ac.kr.
Dong-Young Noh, Email: DYNoh@plaza.snu.ac.kr.
Ida RK Bukholm, Email: IdaBu@ulrik.uio.no.
Monica Nicolau, Email: nicolau@stanford.edu.
Patrick O Brown, Email: Patrick.Brown@stanford.edu.
Robert Tibshirani, Email: tibs@stat.stanford.edu.
References
- Kapp AV, Jeffrey SS, Langerød A, Børresen-Dale AL, Han W, Noh DY, Bukholm IR, Nicolau M, Brown PO, Tibshirani R. Discovery and validation of breast cancer subtypes. BMC Genomics. 2006;7:231. doi: 10.1186/1471-2164-7-231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sørlie T, Tibshirani R, Parker J, Hastie T, Marron JS, Nobel A, Deng S, Johnsen H, Pesich R, Geisler S, Demeter J, Perou CM, Lønning PE, Brown PO, Børresen-Dale AL, Botstein D. Repeated observation of breast tumor subtypes in independent gene expression data sets. Proceedings of the National Academy of Sciences of the United States of America. 2003;100:8418–8423. doi: 10.1073/pnas.0932692100. [DOI] [PMC free article] [PubMed] [Google Scholar]