SiftCell

SiftCell distinguishes cell-containing and cell-free droplets in single-cell RNA sequencing (scRNA-seq) data to improve upstream quality control of transcriptomic profiles.

Key Features:

• Manifold space visualization and randomization: Leverages manifold space visualization combined with randomization techniques to represent the distribution of droplets based on transcriptomic profiles.
• SiftCell-Shuffle: Uses randomization strategies to visualize the distribution of cell-containing versus cell-free droplets within a manifold space.
• SiftCell-Boost: Classifies droplets as cell-containing or cell-free by analyzing manifold space representations using advanced algorithms.
• SiftCell-Mix: Quantifies the contribution of ambient RNAs within each droplet to assess non-cellular RNA influence.

Scientific Applications:

• Upstream quality control for scRNA-seq: Detects and filters cell-free droplets to enhance the integrity of single-cell transcriptomic datasets.
• Ambient RNA assessment: Quantifies per-droplet ambient RNA contribution to characterize contamination from damaged or lysed cells.
• Cross-platform applicability: Applies across various single-cell platforms to perform consistent droplet-level quality assessment.

Methodology:

Combines manifold space visualization with randomization (SiftCell-Shuffle), classification of manifold representations using advanced algorithms (SiftCell-Boost), and per-droplet ambient RNA quantification (SiftCell-Mix).