Figure 1
From: scCAN: single-cell clustering using autoencoder and network fusion
The overall analysis pipeline of scCAN consists of three modules. In the first module (A), we perform data normalization, gene filtering, and latent variables generation using two autoencoders. In the second module (B), we adopt the network fusion-based clustering method to segregate cell types for small data. The third module (C) aims at clustering big data using a combination of the network fusion approach and K nearest neighbors (k-NN) algorithm.
