You will find below short summaries of recent research findings generated by groups at the ICCB. For more details, please see the lab homepages or the corresponding scientific publications.
August 2022 - Schwarz Lab
Copy-number dosage regulates telomere maintenance in neuroblastoma
Telomere maintenance is a key convergent phenotype in neuroblastoma realised through different orthogonal genetic aberrations - TERT rearrangements, TERT upregulation through MYCN amplification, or alternative lengthening of telomeres (ALT). ALT is commonly associated with loss-of-function mutations in the ATRX gene, but a substantial number of ALT-positive tumours show no known genetic cause.
The Schwarz lab in collaboration with the Ohler lab at the MDC, Berlin, has analysed 115 whole-genome and RNA-sequenced neuroblastoma primary patient samples. Systematic analysis of copy-number dosage has revealed dysregulation of histone variants H3 and H2A as potential alternative pathways to ALT.
August 2022 - Schwarz Lab
Inferring cancer evolution with whole-genome doublings
Inferring cancer evolution is highly valuable for understanding how tumours develop and progress. Somatic copy-number alterations are a great source of phylogenetic information but difficult to deal with computationally. So far, only few methods could deal with the complexity of copy-number alterations, and none were able to take whole-genome doubling events into account.
The Schwarz lab has developed MEDICC2, a new algorithm for inferring cancer evolution from somatic copy-number alterations from multi-region or single-cell sequencing. MEDICC2 infers the minimum number of copy-number events - gains and losses of arbitrary length as well as whole-genome doublings - to transform one genome into another. It infers the tree topology explaining the data with the smallest number of events, reconstructs ancestral genomes and extracts individual evolutionary events.
July 2022 - Schwarz Lab
Simulating cancer evolution with one billion cells
In-silico simulations of cancer growth and evolution are useful tools for understanding the fundamental processes shaping cancer genomes. So far, these models did either not take spatial constraints into account and were unable to recapitulate the rich clonal dynamics observed in real-world tumours, or they were computationally too complex to simulate tumours with realistic sizes of one billion cells or more.
The Schwarz lab has developed SMITH, a new simulation framework that combines the classic branching process model of cancer with a new confinement mechanism which limits clonal growth based on the size of the tumour, simulating spatial constraints. SMITH recreates clonal dynamics otherwise only known from lattice-based spatial models while remaining efficient for large tumours.
July 2022 - Schwarz Lab
Reconstructing haplotypes from chromatin conformation data
Chromatin conformation is crucial for precise gene expression regulation and healthy development. Allele-specific differences in the conformation of homologous chromosomes are clinically relevant, but not trivial to obtain. Genome Architecture Mapping (GAM) infers chromatin conformation by sequencing spatially proximal DNA fragments captured in ultra-thin slices of nuclei. Due to the spatial proximity of captured genomic regions, GAM also stores valuable haplotype information.
The Schwarz lab in collaboration with the Pombo lab at the MDC, Berlin, has developed GAMIBHEAR, a novel, graph-based algorithm for haplotype reconstruction from chromatin conformation data. GAMIBHEAR leverages haplotype fidelity of genomic variants captured in GAM data and infers highly accurate and complete, chromosome-spanning haplotypes.
