Chromosomal instability (CIN) is a hallmark of cancer and comprises structural CIN (S-CIN) and numerical or whole chromosomal CIN (W-CIN). Recent work indicated that replication stress (RS), known to contribute to S-CIN, also affects mitotic chromosome segregation, possibly explaining the common co-existence of S-CIN and W-CIN in human cancer. Here, we show that RS-induced increased origin firing is sufficient to trigger W-CIN in human cancer cells. We discovered that overexpression of origin firing genes, including GINS1 and CDC45, correlates with W-CIN in human cancer specimens and causes W-CIN in otherwise chromosomally stable human cells. Furthermore, modulation of the ATR-CDK1-RIF1 axis increases the number of firing origins and leads to W-CIN. Importantly, chromosome missegregation upon additional origin firing is mediated by increased mitotic microtubule growth rates, a mitotic defect prevalent in chromosomally unstable cancer cells. Thus, our study identifies increased replication origin firing as a cancer-relevant trigger for chromosomal instability.

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.

Translatingin vitroresults 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 setF₁score >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.

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 F ₁ 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. Author summary: Cancer researchers perform experiments with cell lines to better understand the biology of cancer and to develop new anti-cancer treatments. A prerequisite to translate promising results from these in vitro experiments to clinical applications is to use the most appropriate cell line for a given tumour or cancer subtype. We present MFmap (model fidelity map), a deep learning technique to integrate cancer genomic data from patients with cell line data. The MFmap neural network compresses complex genomic features from thousands of genes into a small set of features called latent representations. This makes cell line and tumour data comparable and allows cancer researchers to select the best cell line which closely resembles a specific type of tumours or even an individual tumour. By classifying cancer cell lines into subtypes, MFmap offers a new possibility to predict the effect of therapeutic compounds in a particular tumour subtype. For the example of an aggressive brain tumour we demonstrate that MFmap can be used to study cell-state transformations during the disease course. In addition, MFmap is a promising machine learning method with potential applications in many other areas of biology and medicine.