Cliff, A Bioinformatics Software Tool To Explore Molecular Differences Between Two Sets Of Cancer Cell Lines

Cancer cell lines are important models for drug tests and viability screens. For the evaluation and understanding of differences between two groups of cell lines, we created a research software application called CLIFF (cell line differences). CLIFF finds differences in mutational pattern or DNA copy number, lists and visualizes the significantly differentially expressed genes or proteins or performs Gene Set Enrichment Analysis (GSEA) to name a few. Today we have data from large panels of cancer cell lines consisting of gene-, protein- and microRNA expression, DNA copy number, DNA mutations, methylation, histone H3 modification, and concentration of metabolites, in addition to annotations of the tumor of origin. A number of cell line properties have been derived from these data, like microsatellite-(in)stability (MSS/MSI), mutation pattern classifications, or mutational burden. Many drugs have been tested and genes have been knocked in and out to study changes in molecular processes and proliferation. Even though cell lines in vitro do not represent the growth-properties and -conditions in a multi-cell-type tumor environment very well, they are indispensable cancer models in research and drug discovery. After testing or evaluating various properties of cell lines, like e.g. the viability dependency on short hairpin RNA (shRNA) knockdown or CRISPR-Cas9 knockout, as well as the expression of a specific gene, the presence or absence of a DNA mutation, or any other of the above-mentioned annotations, cell lines can be classified into two classes. In addition, users can upload cell line classifications based on external data, e.g. sensitivity or resistance to drug treatments or growth media formulations. Here are two examples of the application of CLIFF: (1) the association of knockdown or knockout of WRN in colorectal cell lines with microsatellite instability (MSI) could be confirmed. (2) The growth media has an important impact on the dependency on certain gene knockouts. It could be verified, that low concentrations of L-asparagine in the growth medium is associated with increased dependency on the asparagine producing enzyme asparagine synthetase (ASNS). In summary, we integrated various cell line data into CLIFF to provide sophisticated statistical analysis to researchers without deeper bioinformatics knowledge. This supports the understanding of molecular mechanisms of tumor growth, helps defining drug targets for treating cancer within defined patient populations, and accelerates our understanding of the molecular drivers of cancer. Citation Format: Andreas Wernitznig, Jesse J. Lipp, Thomas Zichner, Daniel Gerlach, Markus J. Bauer, Tilman Voss, Andreas Schlattl, Christian Haslinger, Philip G. Montgomery, Mahdi Zamanighomi, William R. Sellers, Norbert Kraut. CLIFF, a bioinformatics software tool to explore molecular differences between two sets of cancer cell lines [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3227.