plant-msubp

This repository provides a simple script to build and run the service on any Docker-enabled host.

This application predicts the subcellular localization of proteins using sequence-based features and machine learning. It supports both single-location and multi-location protein prediction, covering 11 cellular compartments and 3 biologically relevant multi-location classes.

Protein sequences are represented using a combination of composition-based and physicochemical descriptors, including amino acid composition, dipeptide composition, pseudo amino acid composition, and hybrid feature sets to improve prediction accuracy.

---

Requirements

• Docker installed and running
• Permission to run docker
• Port 3838 available

---

Quick Start

Build and run the service:

./build-docker.sh

This will:

• build the Docker image
• start the service container
• restart it automatically unless stopped

---

Networking

The container runs on a Docker network named:

webapps-net

The network is created automatically if it does not exist.

---

Image Tags

Each build creates:

• a timestamped image tag (for reference)
• a stable tag (current) used to run the container

The running container always uses the current tag.

---

Service Details

• Container name: plant-msubp

• Port mapping:

  3838 -> 3838
  

• Restart policy: unless-stopped

---