J-SPACE

J-SPACE simulates spatial cancer evolution and associated sequencing experiments to generate synthetic NGS datasets for benchmarking and analysis of evolutionary and sampling effects.

Key Features:

• Spatial Dynamics Simulation: Models spatial interactions of cancer cells as a continuous-time multi-type birth-death stochastic process on arbitrary graphs with configurable interaction rules.
• Evolutionary Dynamics Modeling: Simulates genomic alterations including single-nucleotide variants and indels under the Infinite Sites Assumption and mutational signatures-based substitution models.
• Optimized Gillespie Algorithm: Employs an optimized Gillespie algorithm to simulate stochastic cell dynamics and mutation propagation efficiently.
• Phylogenetic Model Generation: Generates phylogenetic models from simulated spatial sampling of tumor cells to represent evolutionary relationships among subpopulations.
• Synthetic Read Generation: Produces synthetic reads from NGS platforms using the ART read simulator to create realistic sequencing data.
• Output Formats: Exports data in FASTA, FASTQ, SAM, ALN, and Newick formats.
• Incomplete Sampling and Measurement Error Simulation: Simulates incomplete sampling and experiment-specific measurement-related errors to reflect realistic sequencing noise.
• Algorithmic Scalability and Expressivity: Implements an algorithmic framework that balances scalability and expressivity for complex cancer evolution scenarios.

Scientific Applications:

• Spatial dynamics investigation: Investigates emergent spatial dynamics of cancer subpopulations and their interactions within tumor microenvironments.
• Impact of variability and errors: Assesses how biological variability and measurement-related errors affect sequencing data and downstream inference.
• Benchmarking bioinformatics pipelines: Generates synthetic datasets for evaluating the performance of bioinformatics methods processing NGS data.
• Sampling and error effect analysis: Studies the effects of incomplete sampling and experiment-specific errors on downstream evolutionary and phylogenetic analyses.

Methodology:

Simulations use a continuous-time multi-type birth-death stochastic process on arbitrary graphs, an optimized Gillespie algorithm for event simulation, substitution models including the Infinite Sites Assumption and mutational signatures-based models, ART for synthetic read generation, and subsequent phylogenetic model generation from spatial sampling.