No-Code Track

Three Routes to VFB Data

Choose the right tool for your needs:

Route A: Python (vfb_connect) — Bulk Analysis

The vfb_connect Python API provides fast, programmatic access to VFB data.

Best for: Large-scale analysis, reproducible pipelines, bulk data access, repeatable results

Example:

from vfb_connect import vfb
df = vfb.get_instances("DA1 lPN")  # Returns DataFrame with all instances

Route C: VFB Chat — Quick & Easy

chat.virtualflybrain.org is a guardrailed natural-language interface to VFB — nothing to install or configure. Perfect for quick answers on the go, or when you don’t have access to a more powerful AI.

Best for: Fast lookups, teaching demos, getting started without setup

Example:

“Where do I find DA1 lPN neurons in VFB, and which connectomes have them?”

Route B: MCP Tool — Your LLM + VFB Data

The VFB MCP tool runs inside your own LLM (Claude, etc.) and lets you chain VFB queries with your own reasoning and all your connected data.

Best for: Deep analysis, combining VFB data with your own datasets, leveraging your LLM’s full capabilities

Setup: Follow the MCP setup guide.

Example prompt:

“Search VFB for the neuron type ‘DA1 lPN’ and list every individual neuron across all datasets, with their dataset and VFB ID.”

Tips for Good Answers

• Name the entity precisely — a neuron type, region, gene, or a VFB/FlyBase ID removes ambiguity
• Ask for the identifiers — end with “…and give me the VFB IDs” so you can carry results into vfb_connect for reproducible re-runs
• Say which dataset when it matters (“in FlyWire”, “in the male-CNS”, “in BANC”)
• One step at a time — search, then drill in, then compare

Work Through the Problems

Use the route selector on each problem page to see the MCP and Chat prompts for each research question.