Using VSCode with Tufts HPC

We recommend Visual Studio Code for working with tucca-rna-seq on the Tufts HPC cluster. It gives you a familiar editor, an integrated terminal, and a file browser over the cluster filesystem. Tufts Research Technology documents two supported ways to run it, and that guide is the authoritative source; the notes below just orient you and call out what matters for this workflow.

Tufts HPC Guide: VSCode

Establish a VPN connection

Any off-campus (non-Tufts-network) access requires the Tufts VPN first.

Consider the VPN even on campus

Connecting over Tufts_Secure Wi-Fi without the VPN can cause intermittent drops that interrupt a remote editing session. The VPN is more stable and prompts for Duo less often.

Two supported ways to run VSCode on the cluster

1. OnDemand “VSCode Server” (browser, simplest). Launch it from the Tufts OnDemand portal as an interactive app. It runs on a compute node, so you get a full VSCode in the browser without installing anything locally.
2. Local VSCode + code tunnel. Start an interactive session (srun -p batch --pty bash), module load vscode-cli, run code tunnel, authenticate with a GitHub or Microsoft account, and connect from your local VSCode (Remote-Tunnels extension) or a browser.

Python environments and the tunnel

Per the Tufts guide, “VSCode does not support Python environments to be ported through the remote tunnel.” Activate your conda/miniforge environment in the integrated terminal rather than relying on the editor to pick it up. Either way, end your OnDemand session or code tunnel when you’re done so the compute resources free up for others.

Recommended extensions for this workflow

Once connected, these extensions make editing tucca-rna-seq configs and results more comfortable (install from the Extensions view, Ctrl/Cmd+Shift+X):

• Snakemake Language and Snakefmt: syntax and formatting for the Snakefiles.
• Python and Pylance: for the workflow’s Python scripts.
• R: for the downstream DESeq2 / enrichment R code.
• Rainbow CSV: readable sample sheets and results tables.

For everything else (connection troubleshooting, OnDemand app options, and the current vscode-cli module version), defer to the Tufts HPC VSCode guide.