It’s been a decade since the Human Genome Project finished sequencing our genome, and everyone and their grandmother knows that gene expression controls critical biological processes. Genes encode proteins, and proteins are the building blocks of the human body.
But a single gene can have multiple functions. How? One way is by altering the action of RNA, the intermediary between DNA and finished proteins. When DNA is transcribed into RNA, other proteins in the cell can detect motifs in the RNA and cause it to be modified or destroyed, altering the function of the gene.
A group of U of T researchers led by professors Tim Hughes (Terrence Donnelly Centre for Cellular and Biomolecular Research) and Quaid Morris (CCBR, ECE) and including ECE professor Brendan Frey, have amassed a large collection of 207 RNA motifs. Their work was published in the latest issue of Nature, out July 11.
“The compendium of motifs is more comprehensive than anything generated before and advances the field a leap forward,” said Frey. He is interested in how these motifs and other patterns in DNA and in RNA form a ‘regulatory code’ that dictates how and when genes will be expressed, modified and destroyed, depending on conditions within the cell. “My group is now using the motifs to produce a more accurate regulatory code for alternative splicing.”
Alternative splicing is a process whereby cells cut up and reassemble, or splice, RNA strands. Most human genes get spliced in more than one way, for example, enabling a single gene to be expressed differently in a liver cell versus a brain cell. Using machine learning techniques such as deep learning, Frey’s group has developed a method for predicting RNA splicing patterns, identifying regulatory programs in different tissues, and locating previously unknown regions that control alternative splicing.
What are the applications of this discovery? Frey said that now that a good chunk of RNA motifs are known, he and others can look for mutations in DNA that change those motifs, disrupt the regulatory code, and cause disease. “This work might someday allow biomedical researchers to predict the disease-related effects of genetic mutations and even develop treatments for them, using our regulatory code as a kind of ‘cell simulator’.”
Read the paper on the Nature site:http://www.nature.com/nature/journal/v499/n7457/full/nature12311.html.