Meet SCAVENGE: A Search Engine to Uncover How Genetic Variations Affect Our CellsNews
The Context: Variants in our genes can be important indicators of our risk for developing certain diseases, but there hasn’t been a good way to determine which cell types these genetic changes are acting on. Solving this would help scientists understand how exactly these variants can cause a disease to manifest and which cells to target with new therapies.
The Study: Researchers have built a new computational method, called SCAVENGE, that functions as a search engine to find which cell types are most likely to be affected by disease-causing variants. The tool, inspired by an algorithm for website ranking developed by Google, was pioneered by a team at Boston Children’s Hospital led by NYSCF – Robertson Stem Cell Investigator Vijay Sankaran, MD, PhD. The study appears in Nature Biotechnology.
The Importance: SCAVENGE helps uncover key biological insights into how genetic variants affect our cells, which will ultimately lead to a better understanding of a wide range of diseases and how to treat them.
Scientists have long known that genetic variants can affect many aspects of our biology: everything from our risk of developing Alzheimer’s to how many white blood cells we make. But how exactly do these genetic variants act on our cells to do this? Dr. Sankaran’s team wanted to find out, and turned to new advances in studying single cells in great detail to get to the bottom of this age-old question.
SCAVENGE: A Tool for Taking a Deep Dive into the Genetics of Single Cells
Dr. Sankaran’s team developed a powerful computational tool called SCAVENGE (Single Cell Analysis of Variant Enrichment through Network propagation of GEnomic data) that combines two critical technologies: single cell sequencing (the ability to parse out the genetics and behavior of, you guessed it, single cells) and Genome-Wide Association Studies (studies that look at the genetics of large groups of people – many, many cells – to find genetic variants associated with disease).
This tool uses an algorithm similar to the one Google once employed for ranking websites in order of importance; here ranking individual cells according to their association level with a trait or disease instead. Knowing which cell types are most affected by genetic variants associated with a disease will help scientists figure out how to target them with new treatments.
“We envision that the SCAVENGE framework can enable biological insights, akin to the way that search engines such as Google have accelerated our ability to find relevant information across the vast sea of information on the internet,” the authors say.
Putting SCAVENGE to the Test
The lab started by using SCAVENGE to identify which cell types were most affected by different genetic variants associated with blood cell traits (i.e. distinctive characteristics such as the number of platelets in a drop of blood, or the volume of individual red blood cells). The tool successfully assigned each trait to one or multiple relevant cell types after searching across different populations of blood cells, listing the hits in order of relevance.
They next applied SCAVENGE to blood cells from individuals with COVID-19 and healthy controls to see if the tool could identify which types of cells are most affected by genetic variants associated with an increased risk for severe COVID-19. They found that these variants were having the biggest effect in monocytes and dendritic cells (cells known to be highly impacted in severe COVID-19).
The authors then tested if SCAVENGE could identify which blood cell type is most associated with genetic variants that predispose someone to developing B cell acute lymphoblastic leukemia – the most common childhood cancer. Notably, the precise cells of origin have not been defined in this disease.
Their analysis revealed that acute lymphoblastic leukemia risk variants had the greatest impact on pre-B cells – precursor cells to mature, antibody-producing B cells – suggesting a possible cell of origin for this disease.
The researchers hope this tool will help scientists better understand how genetic variants act on different cell types to alter their function and influence disease risk, in turn leading to new opportunities for treatments.
Variant to function mapping at single-cell resolution through network propagation
Fulong Yu, Liam D. Cato, Chen Weng, L. Alexander Liggett, Soyoung Jeon, Keren Xu, Charleston W. K. Chiang, Joseph L. Wiemels, Jonathan S. Weissman, Adam J. de Smith & Vijay G. Sankaran. Nature Biotechnology. 2022. DOI: https://doi.org/10.1038/s41587-022-01341-y