Directly to content
  1. Publishing |
  2. Search |
  3. Browse |
  4. Recent items rss |
  5. Open Access |
  6. Jur. Issues |
  7. DeutschClear Cookie - decide language by browser settings

High-confidence fusion gene detection in different tumor entities & biomarker discovery in breast cancer

Huang, Zhiqin

PDF, English
Download (9MB) | Terms of use

Citation of documents: Please do not cite the URL that is displayed in your browser location input, instead use the DOI, URN or the persistent URL below, as we can guarantee their long-time accessibility.


Fusion genes play an important role in the tumorigenesis of many cancers. Next generation sequencing (NGS) methods such as RNA-seq provide accurate, high-resolution data, which makes unbiased fusion detection much more feasible. Most fusion detection tools based on RNA-seq data report a great number of candidates (mostly false positives), making it hard to prioritize candidates for validation. I therefore developed confFuse, a scoring algorithm to reliably select high-confidence fusion genes which are likely to be biologically relevant.

Compared with alternative tools based on 96 published RNA-seq samples from six different tumor entities, confFuse dramatically reduces the number of fusion candidates (301 high-confidence from 8083 predicted fusion genes, ~3.7%) and retains high detection accuracy (recovery rate 85.7% of previously validated fusions). Another analysis of 27 unpublished tumors of various origins, results in a recovery rate of ~93% (25/27). Furthermore, a screen of 22 GBM tumors shows 242 high-confidence fusions from 6,018 candidates (~4%), of which ~62% (150/242) were previously validated or harbor supporting reads in DNA-seq. Similarly, in 11 published prostate cancer tumors ~72% high-confidence fusions (17/24 from 849 predictions) have supporting evidence. Validation of 18 high-confidence fusions detected in three primary breast tumor samples resulted in a 100% true positive rate. When applying confFuse on three CLL samples, 15 of 18 candidates were successfully validated. In summary, confFuse can reliably select high-confidence fusion genes that are more likely to be biologically relevant, achieving both high validation rate and high detection accuracy, while reducing the number of candidates to a restricted number for validation.

A genetic analysis of primary and refractory breast cancer tumors identified different aberrations in CNVs, SNVs/Indels and rearrangements. Mutations of microtubule-associated serine-threonine kinase (MAST) and 1-phosphatidylinositol-4,5-bisphosphate phosphodiesterase beta (PLCB) gene family members were only detected in refractory tumors (3/50 and 4/50, respectively). Mutations of members of the calcium channel, voltage-dependent, alpha (CACNA) gene family members, which are involved in the MAPK signalling pathway, are highly prevalent in refractory tumors (24%, 12/50) compared to primary tumors (~2%, 1/46). Rearrangements of CACNA were also identified in one primary and two refractory tumors, and PLCB in three refractory tumors. This suggests that mutations of MAST, CACNA or PLCB gene families may be a novel acquired resistance mechanism in addition to ESR1 mutation.

Hundreds of known or novel fusion genes were identified by confFuse in seven unpublished tumor cohorts, including more than 60 highly reliable fusion proteins in breast cancer. Furthermore, different chromosome-wide enrichments of fusion genes were identified across tumor entities. Overall, a comprehensive landscape of fusion genes in different tumor entities was provided to give an insight for biomarker discovery, especially in breast cancer.

Item Type: Dissertation
Supervisor: Brors, Prof. Dr. Benedikt
Date of thesis defense: 21 March 2016
Date Deposited: 01 Apr 2016 09:31
Date: 2017
Faculties / Institutes: The Faculty of Bio Sciences > Dean's Office of the Faculty of Bio Sciences
Subjects: 004 Data processing Computer science
570 Life sciences
About | FAQ | Contact | Imprint |
OA-LogoDINI certificate 2013Logo der Open-Archives-Initiative