TinGa

TinGa models developmental trajectories in single-cell transcriptomics using the Growing Neural Gas algorithm for trajectory inference of cell developmental dynamics.

Key Features:

• Algorithm: Uses the Growing Neural Gas algorithm to represent and infer developmental trajectories from single-cell transcriptomic data.
• Topology handling: Handles a wide spectrum of trajectory topologies, from simple linear structures to intricate disconnected graphs.
• Benchmarking: Evaluated across 250 diverse datasets, including both synthetic and real data.
• Comparative performance: Demonstrated superior accuracy relative to five state-of-the-art trajectory inference methods in the provided comparisons.
• Computational efficiency: Exhibited the fastest execution times among the compared methods in the reported benchmarks.
• Implementation: Implemented in R.
• Visualization: Results can be visualized using the dynplot package.

Scientific Applications:

• Trajectory inference (TI): Inferring developmental trajectories from single-cell RNA-seq datasets.
• Modeling cell dynamics: Modeling cell developmental dynamics across diverse biological contexts and dataset complexities.
• Large-scale studies: Applicable to large-scale single-cell studies where execution time and scalability are important considerations.

Methodology:

Trajectory inference using the Growing Neural Gas algorithm; benchmarking on 250 synthetic and real datasets with comparisons to five state-of-the-art TI methods; implemented in R and compatible with visualization via dynplot.