Machine Learning Approaches to Intersect Cellular Images and Single-Cell RNA-seq Data

To identify, correlate, and label features that describe where and when a gene is expressed in single-cell imaging data without manual supervision.

This group has published their technology, ImageCCA, which uses sparse canonical correlation analysis (CCA) on histological images to quantify gene expression levels from paired samples. The new approach offers the ability to identify a subset of genes that correlate with morphological features of cells to help integrate image analysis with the molecular detail provided by single-cell RNA-sequencing data.