Chunk #114 — STAR★Methods — Quantification and Statistical Analysis — Manifold learning

Source

Molecular Architecture of the Mouse Nervous System.

Embedded

yes

Text

We then used principal component analysis (PCA) to both reduce noise and to reduce the gene expression space further. Dropping non-significant principal components (Kolmogorov-Smirnov test, p < 0.05) reduced the space to a few tens of dimensions (typically about forty).