If you’re using today’s technologies to profile spatial gene expression, you know it can be difficult to use these data to see how cells interact with their neighbors or how the environment around a tumor impacts the way cancer grows and spreads.

What if you could learn more about each cellular neighborhood, or niche?

Thanks to researchers funded by the Cancer Systems Biology Consortium and Human Tumor Atlas Network, you now have a tool, called covariance environment (COVET), that can help. COVET gives you a more complete representation of gene expression in niche cells, letting you apply powerful existing analysis tools to reveal how cells relate with their neighbors within the tumor environment.

Corresponding author, Dr. Dana Pe’er, of Memorial Sloan Kettering Cancer Center noted that anyone working with today’s single-cell analysis toolkits can use COVET to perform large-scale spatial analysis. “Finding the ideal way to computationally encode a new data type—in this case, spatially resolved gene expression—is the key to unlocking important insights from new technology,” she said.

The research team also developed a deep learning tool, called environmental variation inference (ENVI). ENVI lets you match spatial transcriptomics data with single-cell RNA sequencing data. “Using COVET and ENVI, you can infer the spatial context and full transcriptome of every cell, despite the limited information found in each of the two input data types,” said Dr. Pe’er.