CloneOT: predicting clone-specific perturbation responses with lineage-supervised neural inverse optimal transport

Key points

• Predicting single-cell perturbation responses requires mapping between two unpaired populations of cells measured before and after the perturbation.1
• Optimal transport (OT) provides a principled framework for constructing such mappings by minimizing the transport cost, and neural OT methods (e.g., CellOT) have been employed for this prediction task.1
• Moreover, \CloneOT\ outperforms five baselines on 14 lineage-tracing perturbation scenarios of cancer cells, for both seen and unseen clones.1
• Furthermore, we show that \CloneOT\ reveals the clone-specific transcriptional programs of drug-tolerant persister cells and forecasts the fate of each clone from its cells before treatment.1

Sources (1)

• [1]CloneOT: predicting clone-specific perturbation responses with lineage-supervised neural inverse optimal transport
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 12, 03:02 AM

  Here, we propose \CloneOT, a neural inverse OT method that learns both the transport cost and the transport map from partially lineage-traced single-cell datasets. \CloneOT\ represents the cost as a neural network.

  Predicting single-cell perturbation responses requires mapping between two unpaired populations of cells measured before and after the perturbation.

Extractive summary: sentences quoted from the sources.