Hit the Ground Running

A powerful & flexible toolkit designed for the bioinformatics developer

Start Developing Immediately

Use Watershed's powerful hardware without any changes to your workflow. Your first day on Watershed, you can:

1. Open a terminal and run STAR alignment on some fastq files, pulling data and saving results to an onboard Linux filesystem.

2. Import those results into a Python or R notebook, analyze them with DESeq2, and plot an interactive volcano plot with Plotly

3. Save an HTML report and download it or send it to a colleague.

$ bash speedy_aligner.sh --r1 rawdata/fastq1.fq --r2 rawdata/fastq2.fq > processed/counts_table.csv

###################################

import pydeseq2
import watershed
import polars as pl

counts = pl.read_csv("processed/counts_table.csv")
...

Simple wrapper, serious benefits

When you're ready for the next step, take that STAR alignment, or any intensive scripts, and wrap them in a Watershed Operation. This simple wrapper lets you automatically dispatch your code into a dedicated HPC, unlocks effortless parallelization, and caches and tracks your results. Never lose track of what you did or re-run heavy workloads unnecessarily.

{
    inputs: ["my_fastqs"],
    outputs: ["counts_table"],
    script: "speedy_aligner.py"
}

RunOperation(inputs = [fastq_set1, fastq_set2, fastq_set3])
-----------------------------------------------
> [ws_log] three inputs detected... spinning up 3x worker nodes
> [ws_log] begin parallel execution...
> [ws_log] Complete. Results & logs saved to Data Manager

Turn variables into UI elements

Why stop there? Instead of downloading and sharing an HTML report, Watershed makes UI creation a breeze. No more callbacks or reactive frameworks — you can take any variable in your notebook and turn it into a smart, stateful UI with a single Watershed function. Then, roll up the code with a click of a button and your development notebook just became a user-friendly app.

CPUMemory

from watershed import *
from plotnine import ggplot, aes, geom_bar
import polars

genome = load_genome("hg38")
fastqs = load_sequence("cd47_project_fq")
metadata = load_data("metadata")

alignment = star_solo(genome, fastqs)
deg = edgeR(alignment, metadata)

(ggplot(deg.files("sig_results"), aes("Sample", "fold_enrichment")) +geom_bar())

Go ahead, hand it off

Now, your whole workflow is in a single notebook. Heavy computation is cached and safe from careless re-runs, and exploratory data analysis is performed through a quick UI you built. The last step is to use Watershed's rich project management tools to give your colleagues access to this workflow, and put the power — and responsibility — of their analysis into their own hands.

No, Really - Get Started in Seconds...

Watershed's ethos of building tools that support you but never get in your way mean you can be coding as soon as your project loads.

No-Nonsense Server Launching

Control resources — from a lightweight prototyping sandbox to a GPU-fueled high-RAM powerhouse — and spawn a project using a couple of buttons and sliders.

Work how YOU know best

If you know Python, R, or BASH, you know Watershed. Launch a notebook, a terminal, or an RStudio IDE session with a click of a button and get to work.

Love Your Local? Keep It!

Already have the perfect setup on your local machine? Open an SSH port with a button and get all the Watershed power in the comfort of your own home.

Use Your Favourite Tools

Standard tooling + Internet access means you can install libraries with pip, conda, renv, and more; curl or wget your data; connect to services through their CLI. Make Watershed your environment, right out of the box.