Overview

iscc (in silico cancer center) is a multi-modal tumor evolution data simulator. It grows one selection-driven, spatially-structured tumor, optionally treats it, samples it, and generates single-cell and bulk DNA, RNA, and spatial data, down to sequencing reads. Because every run knows the true clones, mutations, copy numbers, cell states, spatial niches, and lineages, iscc provides a realistic ground truth for benchmarking computational tumor evolution methods.

Main features

• Multi-modal data generation. DNA, RNA, and spatial data are emitted from the same evolving cells, so cross-modal relationships are coherent by construction.
• Ground truth for benchmarking. Every run knows the true clones, copy numbers, mutations, cell states, spatial niches, and lineage.
• Calibrated to real data. Assay parameters can be fit from a pilot dataset, and the defaults were pre-tuned to fit real data.

The pipeline

iscc is a pipeline with a command-line tool associated with the main steps:

Stage	CLI	Does
1. Grow	isccsim	grow (and treat) a tumor; write ground-truth genotypes / CNVs / expression / spatial state
2. Sample	isccsample	biopsy / dissociate the tumor into a sample of cells
3. Assay	isccdata	generate single-cell & bulk DNA/RNA and spatial-transcriptomics data
4. Visualize	isccfig, isccgif	Muller plot, 2D slice, clone tree, and growth animations

Quickstart

isccsim --sim-config config.yaml --steps 2000 --random-seed 0 -o sim_out

The shipped defaults produce a realistic multi-clone tumor, and tumor.diagnose() flags a degenerate tumor after growth and tells you which knob to turn.

Reproducing the landing animation

The animation on the home page is one simulated tumor carried through the whole clinical arc: a primary ductal lesion fills the ducts (DCIS), a subclone breaches into the stroma (IDC), it seeds a metastasis, and a resistant clone relapses after chemotherapy. It is reproducible from the ordinary iscc CLI by using the provided config file:

# 1. run the whole clinical arc (grow -> resect the primary -> chemo -> relapse) and write the
#    compartment trajectory the renderer consumes
isccsim --sim-config configs/landing.yaml -o out

# 2. fully-labelled version (use --splash to drop the annotations as in the home page)
isccgif out --compartment -o landing_full.gif

configs/landing.yaml holds the exact parameters (genome, selection, ductal-field, metastasis and treatment settings) and the timed treatment schedule: (the multi-phase arc, including the surgical resection). Clones are coloured by their dominant selective trait and have the same colour in the grids and the Muller plots:

Colour	Trait
blue	proliferation
orange	duct escape
green	stromal survival
purple	metastasis establishment
red	chemotherapy resistance