Analysis

Analysis are accompanied with detailed instructions. See below.

Identifying aneuploid cells

Analysis	Related figures
Identifying aneuploid cells (link)	Fig. 1; Extended Data Fig. 1.

Cancer-intrinsic archetypes

Analysis	Related figures
Identifying archetypes using scRNA-seq data (link)	Fig. 2; Extended Data Fig. 2
Identifying high cancer cell purity spots of Visium data (link)	Fig. 2; Extended Data Fig. 2
Identifying cell types (especially aneuploid cells) in Visium HD data (link)	Extended Data Fig. 2
Clinical association of archetypes and response groups (link)	Fig. 2
Clinical association of archetypes and overall survival (link)	Fig. 2
Predict archetypes in external cohorts (link)	Fig. 2; Extended Data Fig. 2

Gene expression metaprograms of cancer cells

Analysis	Related figures
Identifying metaprograms (link)	Fig. 3
Determining cellular frequencies of metaprograms (link)	Fig. 3
Inferring cell-cycle phases of cancer cells (link)	Fig. 3i

Cell states of immune and stromal cell types

Analysis	Related figures
Identifying cell states of immune and stromal cell types (link)	Fig. 4
Determining and comparing cell percentages of cell states (link)	Fig. 4

Ecotypes

Analysis	Related figures
Identifying ecotypes (link)	Fig. 5
Comparing ecotypes across patient groups (link)	Fig. 5
Ligand-receptor inference (link)	Extended Data Fig. 8d

Spatial Niches

Analysis	Related figures
Identifying cell types in Xenium data (link)	Fig. 1
Identifying cell states in Xenium data (link)	Fig. 3;5
Identifying spatial niches using Xenium data (link)	Fig. 5

Cell-based classifier of predicting response

Analysis	Related figures
Training and testing the cell-based classifier (link)	Fig. 6a-c

Gene-based classifier of predicting response

Analysis	Related figures
Training the gene-based classifier (link)	Fig. 6d-g
Testing the gene-based classifier using other cohorts (link)	Fig. 6d-g