APBN New Site

APBN Developing Site

Cataloguing Cancer “fingerprints” to Map Cancer Genomes

In a decade-long global effort by world class institutions in Singapore, the UK and USA, has developed the most detailed catalogue of mutational fingerprints found in most types of cancers that could help clarify their developmental history and lead to new prevention and treatment strategies.

A collaborative effort by an international consortium of scientists has identified 81 mutational ‘signatures’ that provide possibilities in uncovering the origins and development of various types of cancer, thereby formulating new strategies for diagnosis and treatment.

The study was published in Nature showing that these signatures can provide a snapshot of a cancer cell’s life cycle and insights to factors related to mutations in the cell’s genetic material.

“Different kinds of DNA copying problems and mutation-causing agents, like tobacco, UV light, and chemotherapeutic drugs, lead to mutations with recognisable fingerprints, which we call mutational signatures,” explained Professor Steven Rozen, a senior author of the study from the Cancer and Stem Cell Biology Programme at Duke-NUS Medical School in Singapore.

The research was part of the Pan-Cancer Project, a massive international effort to establish the most comprehensive map of primary cancer genomes to date, involving more than 1,300 scientists and clinicians from 37 countries. For this particular study, the researchers used machine learning to computationally mine mutation data from almost 24,000 human cancer samples. The large data set allowed them to identify 81 different mutational signatures.

“This study presents a fundamental resource that will tie in with experimental studies to open new doors for understanding the causes of cancer,” said Prof Rozen, who is also Director of Duke-NUS’ Centre for Computational Biology. “This will illuminate new opportunities for cancer prevention or help screen exposed individuals more intensively. The results will also help us understand in more detail how exposure to mutation-causing agents leads to cancer.”

The research was part of the Pan-Cancer Project, a massive international effort to establish the most comprehensive map of primary cancer genomes to date, involving more than 1,300 scientists and clinicians from 37 countries. For this particular study, the researchers used machine learning to computationally mine mutation data from almost 24,000 human cancer samples. The large data set allowed them to identify 81 different mutational signatures.

Prof Mike Stratton, a senior author of the study and Director of the Wellcome Sanger Institute in the UK, said, “Using our detailed catalogue of the range of mutational signatures in cancer DNA, researchers worldwide will now be able to investigate which chemicals or processes are linked to these signatures. This will increase our understanding of how cancer develops, and discover new causes of cancer, helping to inform public health strategies to prevent cancer.” [APBN]