CREME is an innovative virtual lab that enables researchers to simulate reductions in gene activity. It serves as a groundbreaking tool for discovering and comprehending vital elements of the genome. Ultimately, it holds the potential to empower scientists without access to physical laboratories to make groundbreaking advancements.
Picture yourself sifting through countless genetic mutations, many of which remain enigmatic. Utilizing CRISPR gene-editing technology, a small number of these alterations could offer therapeutic possibilities. Yet, validating these possibilities would require extensive hours of laboratory work. Determining which mutations merit further investigation is both time-consuming and expensive. But what if this could all take place in a virtual environment with the help of artificial intelligence?
CREME is a cutting-edge AI-driven virtual lab developed by Peter Koo, an Assistant Professor at Cold Spring Harbor Laboratory (CSHL), along with his team. This tool enables geneticists to conduct thousands of virtual experiments effortlessly. It empowers scientists to start pinpointing and understanding crucial areas of the genome.
The platform is inspired by CRISPR interference (CRISPRi), a method for altering gene activity using CRISPR. CRISPRi allows biologists to downregulate specific genes in a cell. In contrast, CREME enables researchers to replicate these modifications in a virtual genome and forecast their impact on gene activity. Essentially, it’s akin to an AI-assisted version of CRISPRi.
“Performing CRISPRi in the lab is quite difficult and comes with limitations in terms of perturbations and scale. But because our perturbations are conducted virtually, we can extend the limits. The volume of experiments we’ve executed is unprecedented — hundreds of thousands of perturbation trials,” Koo explains.
His team evaluated CREME against another AI-based genome analysis tool called Enformer. They aimed to understand how Enformer’s algorithm predicts genomic outcomes. Questions of this nature are central to Koo’s research focus.
“We have robust models that are impressive in taking DNA sequences and predicting gene expression. However, our understanding of how these models derive their predictions is lacking. They likely make accurate forecasts due to their grasp of gene regulation principles, yet we remain unaware of their underlying reasoning,” Koo states.
Through CREME, Koo’s team was able to uncover a set of genetic principles that Enformer discovered while examining the genome. This understanding could be immensely beneficial for future drug development. “Recognizing the principles of gene regulation provides more avenues for adjusting gene expression levels in accurate and foreseeable manners,” remarks Koo.
With further refinements, CREME may eventually guide geneticists toward uncovering new therapeutic targets. Most significantly, it has the potential to empower researchers without access to physical labs to make these significant discoveries.
