Publications

What is a Publication?
3 Publications visible to you, out of a total of 3

Abstract (Expand)

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.

Authors: X. Zhang, M. Kschischo

Date Published: 16th Dec 2021

Publication Type: Journal

Abstract (Expand)

Chromosome loss that results in monosomy is detrimental to viability, yet it is frequently observed in cancers. How cancers survive with monosomy is unknown. Using p53-deficient monosomic cell lines, we find that chromosome loss impairs proliferation and genomic stability. Transcriptome and proteome analysis demonstrates reduced expression of genes encoded on the monosomes, which is partially compensated in some cases. Monosomy also induces global changes in gene expression. Pathway enrichment analysis reveals that genes involved in ribosome biogenesis and translation are downregulated in all monosomic cells analyzed. Consistently, monosomies display defects in protein synthesis and ribosome assembly. We further show that monosomies are incompatible with p53 expression, likely due to defects in ribosome biogenesis. Accordingly, impaired ribosome biogenesis and p53 inactivation are associated with monosomy in cancer. Our systematic study of monosomy in human cells explains why monosomy is so detrimental and reveals the importance of p53 for monosomy occurrence in cancer.

Authors: N. K. Chunduri, P. Menges, X. Zhang, A. Wieland, V. L. Gotsmann, B. R. Mardin, C. Buccitelli, J. O. Korbel, F. Willmund, M. Kschischo, M. Raeschle, Z. Storchova

Date Published: 22nd Sep 2021

Publication Type: Journal

Abstract (Expand)

Whole chromosome instability (W-CIN) is a hallmark of human cancer and contributes to the evolvement of aneuploidy. W-CIN can be induced by abnormally increased microtubule plus end assembly rates during mitosis leading to the generation of lagging chromosomes during anaphase as a major form of mitotic errors in human cancer cells. Here, we show that loss of the tumor suppressor genes TP53 and TP73 can trigger increased mitotic microtubule assembly rates, lagging chromosomes, and W-CIN. CDKN1A, encoding for the CDK inhibitor p21(CIP1), represents a critical target gene of p53/p73. Loss of p21(CIP1) unleashes CDK1 activity which causes W-CIN in otherwise chromosomally stable cancer cells. Consequently, induction of CDK1 is sufficient to induce abnormal microtubule assembly rates and W-CIN. Vice versa, partial inhibition of CDK1 activity in chromosomally unstable cancer cells corrects abnormal microtubule behavior and suppresses W-CIN. Thus, our study shows that the p53/p73 - p21(CIP1) tumor suppressor axis, whose loss is associated with W-CIN in human cancer, safeguards against chromosome missegregation and aneuploidy by preventing abnormally increased CDK1 activity.

Authors: A. K. Schmidt, K. Pudelko, J. E. Boekenkamp, K. Berger, M. Kschischo, H. Bastians

Date Published: 11th Nov 2020

Publication Type: Journal

Powered by
(v.1.17.1)
Copyright © 2008 - 2025 The University of Manchester and HITS gGmbH