The cell lines that researchers routinely turn to when studying ovarian cancer are not the best genetic match available.

Nikolaus Schultz and his team at the Memorial Sloan-Kettering Cancer Center in New York screened 47 cell lines used to model ovarian cancer. The researchers compared the models with samples of human ovarian cancer in terms of gene expression, genetic duplications and mutations. Many frequently used lines were quite different from the most common and lethal cancer types. For example, two cell lines most often used as tumour models did not contain mutations in a gene that is typically altered in ovarian cancers. Less-used lines were deemed better matches. The analysis will help researchers to choose the most-relevant lines for testing potential cancer drugs, the authors say, and the approach could be used for other tumour types.

Nature Commun. 4, 2126 (2013)