Automorph / Upload & Configure
0 analyses today
Sign In
System Ready
Graph Input Configuration
Edge List

Drop edge-list file here or click to browse

Supports .txt, .csv, .edgelist
Sample Dataset Library
Instant Test
Quick Start
1. Choose Domain
Select your research area from the left sidebar to tailor AI insights.
2. Upload Graph
Upload an edge-list file. Format: #nodes edges followed by pairs.
Agent System Status
Computation Engine Nauty 2.9.3
Compute Latency --
Intelligence Layer Nemotron 3 Super
System responding normally.
Agent System Core
00:00.000
Network Topology
AI Domain Insights
cybersecurity
Insights will appear here after analysis.
Chat Agent
Chat Agent
Analysis context loaded. I can discuss topological properties, group structure, and domain-specific implications. What would you like to know?

Getting Started with Automorph

Learn how to prepare your graph files, run analyses, and explore AI-powered symmetry insights across 6 scientific domains.

1 Graph File Format

Automorph accepts plain text files (.txt or .edgelist) containing an edge list representation of your graph. The format is:

Format Specification graph.txt
# Line 1: number_of_nodes  number_of_edges
5 6
# Subsequent lines: source_node  target_node (one edge per line)
0 1
0 2
1 2
1 3
2 4
3 4
Rules
  • First line must have node count and edge count
  • Nodes are zero-indexed integers (0, 1, 2...)
  • One edge per line, space-separated
  • Undirected graph (edge 0 1 = edge 1 0)
  • Lines starting with # are ignored
Common Mistakes
  • Starting node IDs at 1 instead of 0
  • Missing the header line (node/edge counts)
  • Using commas instead of spaces
  • Including self-loops (e.g. 0 0)
  • Duplicate edges (same pair listed twice)

2 Step-by-Step Walkthrough

1
Select a Domain
Click a domain in the left sidebar (Cybersecurity, PPI Networks, Crystallography, Finance, or Social Networks). This determines the type of AI analysis you'll receive.
2
Prepare Your Graph File
Create a .txt file with the format described above: first line has node and edge counts, followed by one edge per line (space-separated node pairs).
3
Upload & Run Analysis
Go to Upload & Configure, click the file upload area or drag your file in, then press Run Analysis. Watch the Agent System Core terminal as it processes your graph in real time.
4
Review Results
After analysis, you'll see statistics (vertices, edges, group order, orbits), an interactive graph visualization with orbit coloring, and the raw Nauty engine output in the terminal.
5
Get AI Insights & Chat
Click Insights for AI-generated analysis. Then use the Chat Agent to ask follow-up questions about your graph's structure, orbits, and domain implications.

3 Supported Domains

Cybersecurity
Analyze network topologies for symmetric attack surfaces, redundant pathways, and vulnerability equivalence classes. Vertex orbits reveal nodes with identical risk profiles.
PPI Networks
Map protein-protein interactions to find structurally equivalent proteins, functional redundancy, and drug target candidates. Upload an optional protein mapping file for named results.
Crystallography
Identify point group symmetry, Wyckoff positions, and crystallographically equivalent atoms. Optionally provide a molecular formula for enhanced analysis.
Finance Research
Evaluate financial network topology for systemic risk. Orbits reveal institutions with equivalent contagion risk profiles and too-connected-to-fail patterns.
Social Networks
Discover structurally equivalent positions, role equivalence, bridge nodes, and community structure. Understand influence propagation and network resilience.

4 Platform Features

Automorphism Group Computation
Computes the full automorphism group of your graph, including generators, group order, vertex orbits (sets of structurally equivalent nodes), and edge orbits (sets of equivalent edges).
Interactive Graph Visualization
Color-coded interactive graph rendering using vis-network. Nodes in the same orbit share the same color, making structural equivalence visually obvious. Drag, zoom, and explore.
AI-Powered Domain Insights (Nemotron 3 Super)
Streaming AI analysis powered by Nemotron 3 Super that interprets your graph's symmetry properties through the lens of your chosen domain. Insights stream in real-time, token by token.
Chat Agent
An interactive AI chat assistant that knows your graph's full analysis context. Ask follow-up questions about specific orbits, generators, structural properties, or domain-specific implications.
Agent System Core Terminal
Watch the analysis pipeline execute step-by-step in a live terminal. See reasoning steps, tool invocations, observations, and system actions as the agent processes your graph.
PPI Protein Mapping
For PPI Networks, upload an optional mapping file (vertex ID to protein name, one per line as 0 TP53) to see results with actual protein names instead of node numbers.

5 Example Graph Files

Copy any of these examples into a .txt file to try them out. Click to copy.

Triangle (K3)
3 3
0 1
0 2
1 2
Petersen Graph
10 15
0 1
0 4
0 5
1 2
1 6
...
Cube Graph (Q3)
8 12
0 1
0 2
0 4
1 3
1 5
...
Star Graph (S5)
6 5
0 1
0 2
0 3
0 4
0 5
Ready to analyze?
Head to Upload & Configure to run your first analysis.