Criteria for inclusion in the cancer-gene census

列入标准的癌症基因普查

We have exclusively listed genes in which mutations that are causally implicated in oncogenesis have been reported (see supplementary information S1 (table)). Most cancer genes have been identified and initially reported on the basis of genetic evidence (that is, the presence of somatic or germline mutations) and without biological information supporting the oncogenic effects of the mutations. The underlying rationale for interpreting a mutated gene as causal in cancer development is that the number and pattern of mutations in the gene are highly unlikely to be attributable to chance. So, in the absence of alternative plausible explanations, the mutations are likely to have been selected because they confer a growth advantage on the cell population from which the cancer has developed. There often exists additional biological information supporting the role of a mutated gene in oncogenesis. However, if the genetic data is convincing (see below), we have regarded it as sufficient on its own for inclusion of the gene in the census (Box 1).

Genes have been included in the census if there exist at least two independent reports showing mutations in primary patient material. In considering somatic mutations for inclusion in the census, we included only genes for which there was evidence of the somatic origin of at least a subset of mutations, based on analysis of normal tissue from the same individuals. In practice, this is most important for the interpretation of putative somatic base substitutions and small insertions/deletions, because these are also common as germline sequence variants. By contrast, translocations and copy-number changes are relatively uncommon as germline variants. Therefore, we have included some cancer genes that are altered by translocation or copy-number change for which definitive evidence of the somatic origin of their mutations is unavailable.

The census does not include genes that are only expressed at altered levels in cancer cells, without any mutation in DNA. Although alterations in expression might be part of the reconfiguration of cell biology following the acquisition of a mutation in a cancer gene, they are usually not the primary change. Indeed, in some cases they might be consequences of the neoplastic phenotype rather than determinants of it.

The list also does not include genes in which changes in methylation of CpG dinucleotides within promoter regions have been the only reported abnormalities. Many such alterations are clearly associated with altered transcriptional regulation and, in some cases (for example, in the promoters of CDKN2A or MLH1), are likely to cause neoplastic transformation^{6, 7}. Methylation of promoter regions is, however, relatively plastic. It can change in association with age or in vitro culture, is widespread in some cancer genomes and can occur in CpG ISLANDS without any known consequences for transcriptional regulation⁸. It is therefore problematic to identify genes in which alterations in promoter methyation, without mutation, cause neoplastic change, so this form of genetic modification has not been included in our database.

It is likely that some cancer genes have inadvertently been missed, and opinions might differ about some genes that have been excluded and included. We are keen to be informed of these omissions and controversial cases in order to evaluate their status and will regularly update this list on the Wellcome Trust Sanger Institute web site.