Translating in vitro results from experiments with cancer cell lines to clinical applications requires the selection of appropriate cell line models. Here we present MFmap (model fidelity map), a machine learning model to simultaneously predict the cancer subtype of a cell line and its similarity to an individual tumour sample. The MFmap is a semi-supervised generative model, which compresses high dimensional gene expression, copy number variation and mutation data into cancer subtype informed low dimensional latent representations. The accuracy (test set F1 score >90%) of the MFmap subtype prediction is validated in ten different cancer datasets. We use breast cancer and glioblastoma cohorts as examples to show how subtype specific drug sensitivity can be translated to individual tumour samples. The low dimensional latent representations extracted by MFmap explain known and novel subtype specific features and enable the analysis of cell-state transformations between different subtypes. From a methodological perspective, we report that MFmap is a semi-supervised method which simultaneously achieves good generative and predictive performance and thus opens opportunities in other areas of computational biology.

We report on the development of a GEM-based TPC prototype for the PANDA experiment. The design and requirements of this device will be illustrated, with particular emphasis on the properties of the recently tested GEM-detector, the characterization of the read-out electronics and the development of the tracking software that allows to evaluate the GEM-TPC data.

A large proportion of tumours is characterised by numerical or structural chromosomal instability (CIN), defined as an increased rate of gaining or losing whole chromosomes (W-CIN) or of accumulating structural aberrations (S-CIN). Both W-CIN and S-CIN are associated with tumourigenesis, cancer progression, treatment resistance and clinical outcome. Although W-CIN and S-CIN can co-occur, they are initiated by different molecular events. By analysing tumour genomic data from 33 cancer types, we show that the majority of tumours with high levels of W-CIN underwent whole genome doubling, whereas S-CIN levels are strongly associated with homologous recombination deficiency. Both CIN phenotypes are prognostic in several cancer types. Most drugs are less efficient in high-CIN cell lines, but we also report compounds and drugs which should be investigated as targets for W-CIN or S-CIN. By analysing associations between CIN and bio-molecular entities with pathway and gene expression levels, we complement gene signatures of CIN and report that the drug resistance gene CKS1B is strongly associated with S-CIN. Finally, we propose a potential copy number-dependent mechanism to activate the PI3K pathway in high-S-CIN tumours.