Tufts HPC Quick Start Guide

Welcome to the Tufts-specific documentation for the tucca-rna-seq workflow! These guides provide essential information for using the workflow on the Tufts High-Performance Computing (HPC) cluster.

Start Here

This guide provides Tufts-specific quick start information. For complete installation and setup instructions, see the main Installation Guide first.

This guide covers the essential Tufts-specific steps to get you running the tucca-rna-seq workflow on the Tufts HPC cluster quickly.

---

Prerequisites

1. Tufts HPC Account

You need a Tufts Cluster Account to access the HPC cluster:

Request an Account

Requirements:

• Valid Tufts Username and Password
• Faculty or researcher sponsorship (for guest/student accounts)
• Approval process takes multiple business days

TTS Research Technology Guides - HPC Account Request

---

Quick Setup Steps

1. Connect to Tufts HPC

SSH Access (Recommended):

ssh your_utln@login-prod.pax.tufts.edu

Off-campus (any non-Tufts network) requires the Tufts VPN first. For the authoritative connection details, see the Tufts HPC command-line access guide.

VSCode Remote Development (Highly Recommended):

Setting Up Tufts HPC Access in VSCode

2. Get Off the Login Node

Never run compute on a login node. Drop onto a compute node with an interactive session on the default batch partition first:

# Request an interactive compute node (batch partition, 2 h, 1 core)
srun -p batch --time=02:00:00 --pty bash

The cluster’s partitions are batch (CPU, default), gpu, and preempt, each capped at 2 days — see the Tufts partitions guide.

3. Load Required Modules

# Purge all currently loaded modules
module purge

# Load required modules for tucca-rna-seq
module load snakemake/8.27.1
module load singularity/3.8.4
module load miniforge/24.11.2-py312

Module versions drift

The versions above are examples. Run module av snakemake (and likewise for singularity / miniforge) to confirm what’s currently installed, then pin the versions your run actually uses. See the Tufts modules guide.

4. Navigate to Lab Workspace

# For Kaplan Lab users
cd /cluster/tufts/kaplanlab/your_utln

# For other labs
cd /cluster/tufts/your_lab/your_utln

# Create project directory
mkdir my_rnaseq_project
cd my_rnaseq_project

Set Up a Shared Software Cache!

Your home directory on the cluster has a strict storage quota. To avoid errors and speed up future analyses, we strongly recommend setting up a permanent, shared cache for software environments in your lab’s shared workspace before you run the workflow for the first time.

See the HPC Best Practices Guide for detailed instructions.

5. Clone and Setup Workflow

For detailed instructions on cloning the repository and setting up the workflow’s dependencies, please follow the main Installation Guide.

After cloning, you will need to configure the workflow for your specific analysis. This includes editing the main config.yaml file and associated sample sheets, as well as customizing the slurm execution profile to add your email for job notifications.

For detailed instructions, please refer to the main Configuration Guide.

---

Tufts-Specific Resources

Documentation

• Tufts HPC Home Page
• Tufts HPC Cluster Documentation
• HPC Carpentry Lessons

Support

• TTS Research Technology: tts-research@tufts.edu
• Workflow Team: benjamin.bromberg@tufts.edu
• GitHub Issues: Report problems
• HPC User Guide: https://rtguides.it.tufts.edu/hpc/

---

Next Steps

After completing this quick setup:

1. Follow the main Installation Guide for complete workflow setup
2. Configure your analysis using the Configuration Guide
3. Run your workflow following the Running Guide

---

For complete documentation, always refer to the main guides first, then use this quick start for Tufts-specific details.