> ## Documentation Index
> Fetch the complete documentation index at: https://docs.scmcphub.org/llms.txt
> Use this file to discover all available pages before exploring further.

# infercnv MCP Server Tools

Here is doc for infercnv-mcp, but you can refer to infercnvpy docs at [https://infercnvpy.readthedocs.io/](https://infercnvpy.readthedocs.io/)

## Core Analysis Module

* `infercnv`: Perform Copy Number Variation (CNV) analysis by averaging gene expression over genomic regions
* `cnv_score`: Calculate CNV scores for individual cells to quantify copy number variation

## Dimensionality Reduction Module

* `pca`: Perform Principal Component Analysis on CNV data with default parameters optimized for CNV analysis
* `tsne`: Generate t-SNE visualization of CNV data using CNV-specific neighborhood information
* `umap`: Create UMAP projection of CNV data utilizing CNV-specific neighborhood relationships

## Clustering Module

* `leiden`: Perform Leiden clustering on CNV data using CNV-specific neighborhood information
* `louvain`: Apply Louvain clustering algorithm to CNV data using CNV-specific neighborhood relationships

## Data Requirements

The input AnnData object must contain the following columns in `adata.var`:

* `chromosome`: Chromosomal location of genes
* `start`: Start position of genes
* `end`: End position of genes

These columns can be added using the `load_gene_position` function before running CNV analysis.

## Output

The analysis generates two main outputs:

1. CNV scores stored in the original AnnData object
2. A separate CNV-specific AnnData object containing the CNV analysis results

All visualizations and clustering results are stored with CNV-specific keys (e.g., `cnv_pca`, `cnv_umap`, `cnv_leiden`) to distinguish them from other analyses.