Panomics Documentation
  • Overview
  • Projects
    • Project Structure
    • Project Visibility
    • Project Access
  • Samples
    • Sample Lifecycle
    • Sample Ingestion
    • Sample Metadata
  • Analyses
    • Create an Analysis
    • Delete an Analysis
    • Change Analysis Status
    • Publish an Analysis
    • View Analysis Audit Trail
    • Analysis Workbench
      • Quality Control
      • Gene Expression
      • Compute Statistics
      • Sample/Cell Explorer
        • Normalization
        • UMAP
      • Cell Composition
      • Marker Genes
      • Differential Gene Expression
      • Differential Pathway Activity
      • WGCNA
      • Gene Set Enrichment
    • Data Slices
  • OmicsBrowser
  • Gene Signature Similarity
  • Compute Runtimes
    • Codeless Compute
    • Code-based Compute
  • API Keys
  • Administration
    • Roles
    • Users
    • Ontologies
    • Metadata Fields
    • Collections
    • Gene Set Collections
    • Organization Settings
  • Release Notes
  • Technical Support
  • Legal
Powered by GitBook
On this page
  • Runs batch effect correction algorithm (optional)
  • Runs PCA
  • Computes neighborhood graph
  • Computes UMAP
  1. Analyses
  2. Analysis Workbench
  3. Sample/Cell Explorer

UMAP

PreviousNormalizationNextCell Composition

Last updated 7 months ago

The UMAP command in Panomics performs multiple processing steps. A runtime is required.

1

Runs batch effect correction algorithm (optional)

Algorithm
Microarray
Bulk RNA-seq w/ normalized counts
Bulk RNA-seq w/ raw counts
Single Cell RNA-seq

Limma

Combat

CombatSeq

Harmony *

scVI

2

Runs PCA

3

Computes neighborhood graph

Uses 15 neighbors and the PCA representation based on batch effect correction algorithm choice.

4

Computes UMAP

* The Harmony batch effect correction algorithm operates on the PCA space, so PCA is computed first and then replaced by the harmonized PCA.

In the following example, we perform normalization, then UMAP with no batch correction. Coloring the embedding by the diagnosis and sample name observations reveals strong batch effects. We then run UMAP again with Harmony batch effect correction.

✅
✅
❌
❌
✅
✅
❌
❌
❌
❌
✅
❌
❌
❌
✅
❌
❌
❌
✅
Standard Compute
Normalize & UMAP