Researchers at the University of Illinois Urbana-Champaign have developed a new modelling approach, the pGENMi model, to better understand how tumours become aggressive. The researchers pooled the data from their own experiments as well as publicly available data to develop the model, which was based on a simpler 2018 model that investigated regulators of cancer drug resistance. In their paper, ‘An integrated multi-omics approach to identify regulatory mechanisms in cancer metastatic processes’, published Genome Biology, they specifically focused on transcription factors, which are proteins that control gene expression by binding to the DNA.
"Researchers have identified several cellular pathways that change when a tumour becomes aggressive. However, it is difficult to understand how they affect the tumour," said Steven Offer, an assistant professor of molecular pharmacology and experimental therapeutics at Mayo Clinic, Minnesota. "We wanted to develop a simple system that can model how cancer cells form an aggressive tumour."
However, the target areas can be hidden depending on the DNA organisation. By studying their availability, the researchers can predict which transcription factors and targets are important.
"We can easily see how many transcription factors are there in the cancer cell. This model allowed us to see whether the target areas they bind to are available or not," explained Offer.
The researchers also tested the predictions of the model using human cancer cell lines. They looked at the transcription factors that were identified and showed that they were involved in increasing the aggressiveness of the colorectal cancer cells.
"The advantage of the model is that it can integrate different types of experimental data, which is not an easy task. It gave us a list of transcription factors, ranked based on their relevance to colorectal cancer aggressiveness," said Saba Ghaffari, a PhD student in the Sinha lab. The model is so adaptable that it was able to analyse the binding of transcription factors in other types of cells as well.
"Without the model, it would have been expensive and time consuming for us to analyse the transcription factors in all these different cell lines," added Offer. "We can now use this data to improve cancer care. The more information we have about these factors, the more disruptions we can create to interfere with the process and improve treatments."
The researchers are hoping to improve the model further to make it more sensitive.
"Although we binarised the data, the effects of these transcription factors are continuously changing. We also assumed that all the genes work independently of each other; in reality they work together," explained Ghaffari.
"Increasingly, these technologies provide us complementary views of cellular changes during disease progression. Ghaffari's work provides us with a general-purpose recipe to combine those different views into one meaningful whole, giving us more that any one view can," said Saurabh Sinha (BSD/CABBI/GNDP/GSP), a professor of computer science. "This is just the beginning. We are looking at it as a blueprint for many more analyses in the future, tackling different biological challenges."
To view this paper, please click here