Kilosort Docker

Containerized Kilosort 2 spike-sorting environment for reproducible GPU-based electrophysiology data analysis.
This image provides a fully configured MATLAB + CUDA so you can run Kilosort 2 locally or on a cluster without manual setup.

---

What It Does

• Runs Kilosort 2 inside a Docker container with NVIDIA GPU acceleration
• Supports input/output volume mounting for raw data and results
• Compatible with Phy for manual curation
• Optional Kubernetes (k8s/) manifests for large-scale batch jobs

---

File Structure

docker/ → Dockerfiles for Kilosort2 builds
matlab/ → Precompiled MATLAB runtime for KS2
src/ → KS2 scripts, configs, and utilities

---

Quick Start

# Build the image
docker build -f docker/Dockerfile.kilosort2 -t kilosort2:latest .

# Run Kilosort2 on a dataset
docker run --rm --gpus all \
  -v /path/to/data:/input \
  -v /path/to/output:/output \
  kilosort2:latest \
  matlab -batch "run('/opt/kilosort/run_kilosort2.m')"