Single-cell RNA sequencing (scRNA-seq) is an emerging technology that can assess the function of an individual cell and cell-to-cell variability at the single cell level in an unbiased manner. Dimensionality reduction is an essential first step in downstream analysis of the scRNA-seq data. However, the scRNA-seq data are challenging for traditional methods due to their high dimensional measurements as well as an abundance of dropout events (that is, zero expression measurements).
To overcome these difficulties, University of Washington researchers have developed DR-A (Dimensionality Reduction with Adversarial variational autoencoder), a data-driven approach to fulfill the task of dimensionality reduction. DR-A leverages a novel adversarial variational autoencoder-based framework, a variant of generative adversarial networks. DR-A is well-suited for unsupervised learning tasks for the scRNA-seq data, where labels for cell types are costly and often impossible to acquire. Compared with existing methods, DR-A is able to provide a more accurate low dimensional representation of the scRNA-seq data. The researchers illustrate this by utilizing DR-A for clustering of scRNA-seq data.
The novel architecture of an Adversarial Variational AutoEncoder
with Dual Matching (AVAE-DM)
An autoencoder (that is, a deep encoder and a deep decoder) reconstructs the scRNA-seq data from a latent code vector z. The first discriminator network D1 is trained to discriminatively predict whether a sample arises from a sampled distribution or from the latent distribution of the autoencoder. The second discriminator D2 is trained to discriminatively predict whether the scRNA-seq data is real or fake