Logo

Quickstart Guide

The ResearchBox Assistant helps you configure a cloud virtual machine with the research application software, machine learning model, or bioinformatics pipeline of your choice - resulting in a full graphical desktop rendered inside your browser. It can then help you scale that machine to hundreds of cores and hundreds of GB of memory, and back down again.

You can modify the the research software or source code, or build your own pipeline with it using familiar tools like vscode, docker, singularity, bash scripts, bioconductor and anything else you want to install.

1
Create an account
Sign up with your institute or work email address for instant access, otherwise you will be on a waitlist.
2
Ask the assistant to run a researchbox
"run a molecular docking researchbox"
The Assistant will search for molecular docking packages and ask you to select one.
3
Optionally select the cloud machine type to run it on.
The assistant will offer to customize the cloud machine on which to run your package. 
"show me cloud machines with 16 cores and 32 GB ram"
4
Use your ResearchBox
You now have a cloud instance fully configured research environment with the selected package and also development tools like vscode code and docker. You can run the code, modify it, even submit patches to the github repo. And develop your own pipelines with it. Ask the AI assistant to assist you with running the model. 
"how can I run the molecular docking benchmark mentioned in this package"
5
Resize your ResearchBox
Need more cores or more memory or a more powerful GPU? Ask the assistant to resize your box. 
"resize my box I want an nvidia GPU that costs less than 2 usd per hour"
6
Stop and Restart your ResearchBox
You do not get charged for cpu time for stopped boxes. You use the "My Boxes" dashboard to restart your box and open the streaming desktop exactly as you left it.
7
Publish your modifications as a package
Make modifcations to your box, maybe add some code and data to run an expermient. Then ask the assistant to publish it as a researchbox. Now other researchers can run a box from that package and precisely reproduce your research. 
"publish my box as a package"