Fig. 1
From: ENGEP: advancing spatial transcriptomics with accurate unmeasured gene expression prediction
Framework of ENGEP. a The input of ENGEP. ENGEP takes in a spatial transcriptomics dataset and multiple scRNA-seq datasets that profile the same or similar tissue as the spatial dataset. b Generating base results. ENGEP first partitions each substantial reference dataset into smaller sub-reference datasets. Then, base results are generated by ENGEP for each sub-reference and query dataset pair. It utilizes k-nearest-neighbor (k-NN) regression with ten different similarity measures and four different values of k (number of neighbors) to generate these base predictions. c Combining base results. The base results are combined by ENGEP using a weighted average ensemble strategy to produce the final prediction