Welcome to CellDyc

CellDyc is a novel semi-supervised learning framework designed to reconstruct transcriptomic velocities and recover intrinsic “gene-embedded time” by leveraging experimental time-point supervision.

Asynchronous temporal signals

CellDyc

Projection of transcriptomic velocities

Key Highlights:

• Data-Driven Velocity Inference: Provides a robust, data-centric solution for calculating transcriptomic velocities.

• Intrinsic Time Recovery: Accurately recovers the intrinsic “gene-embedded time” from the transcriptome.

• Flexible Supervision: Capable of handling sparse and noisy supervision signals effectively.

• Seamless Integration: Fully compatible with existing computational biology infrastructure and workflows.