Skip to content

Latest commit

 

History

History
62 lines (48 loc) · 2.18 KB

File metadata and controls

62 lines (48 loc) · 2.18 KB

Similarity Network Fusion - Usage Guide

Overview

SNF integrates multiple data types by constructing patient similarity networks for each omics layer and fusing them into a single network that captures shared and complementary information.

Prerequisites

install.packages('SNFtool')
pip install snfpy

Quick Start

Tell your AI agent what you want to do:

  • "Identify patient subtypes by fusing my RNA-seq and methylation data"
  • "Build a similarity network from multiple omics to stratify patients"

Example Prompts

Patient Stratification

"Use SNF to fuse my expression and CNV data to identify cancer subtypes"

"Find patient clusters by integrating RNA, protein, and methylation similarity networks"

Survival Analysis

"Identify subtypes from multi-omics data and test for survival differences"

"Stratify patients using SNF and generate Kaplan-Meier curves per cluster"

Parameter Tuning

"Run SNF with different K values to check cluster stability"

"Estimate the optimal number of clusters from my fused similarity network"

What the Agent Will Do

  1. Load multi-omics data matrices with matched samples
  2. Compute pairwise distance matrices per omics layer
  3. Convert distances to affinity matrices using Gaussian kernel
  4. Fuse networks using iterative SNF algorithm
  5. Estimate optimal cluster number
  6. Perform spectral clustering on fused network
  7. Validate clusters (silhouette, survival, known labels)

Key Parameters

Parameter Description Typical Range
K Number of neighbors for affinity 10-30
alpha Kernel width scaling 0.3-0.8
t Fusion iterations 10-20

Tips

  • Normalize each omics separately (z-score or quantile) before computing distances
  • Filter low-variance features to improve signal
  • All samples must be present across all omics; impute or remove missing
  • Use estimateNumberOfClustersGivenGraph() to choose cluster number
  • Validate clusters with survival analysis if clinical data available
  • Check cluster stability by varying K and alpha parameters

References