Universal Latent Variable Model of Cellular State
Train predictive models that can probabilistically infer any missing
omics data from any available measurements by learning a broadly useful
latent space. Such a universal representation of cellular state could become
the foundation for many predictive and diagnostic applications.