GSoC 2022 Project Idea 22.1: GNNs as Developmental Networks

Biological development features many different types of networks: neural connectomes, gene regulatory networks, interactome networks, and anatomical networks. Using cell tracking and high-resolution microscopy, we can reconstruct the origins of these networks in the early embryo. Building on our group’s past work in deep learning and pre-trained models, we look to apply graph neural networks (GNNs) to developmental biological analysis.

We seek to create graph embeddings that resemble actual biological networks found throughout development. Potential activities include growing graph embeddings using biological rules, differentiation of nodes in the network, and GNNs that generate different types of movement output based on movement seen in microscopy movies. The goal is to create a library of GNNs that can simulate developmental processes by analyzing time-series microscopy data.

DevoWorm is an interdisciplinary group engaged in both computational and biological data analysis. We have weekly meetings on, and are a part of the OpenWorm Foundation. You may also have the chance to work with our DevoLearn (open-source pre-trained deep learning) software, in addition to adding your contributions to the DevoWorm AI library.

What can I do before GSoC?
You can ask one of the mentors to direct you to the data source and you can start working on it. Please feel free to join the OpenWorm Slack or attend our meetings to raise questions/discussions regarding your approach to the problem.

DevoWorm website: link

DevoLearn (preprint): link

DevoWorm AI: link

PyTorch/Tensorflow (PyTorch will be preferred because all our other models are on that framework already) Wrangling with video data Building a simple GUI on top of the model to run it on local systems (on Linux/windows/macOS). Basic knowledge of biology and complex networks theory would be helpful.

Planned Effort: 175 hours Mentors: Bradly Alicea (, TBA



I found the project description very interesting and would like to work on this. I have a decent experience working with Pytorch and contributing to research projects.
What would be a good way to get started working towards my proposal ?


Hi again! I had conveyed my interest in another project on developing a collective cognition model for AI ethics a week or so back. I came across this project at the end of the GSoC ideas list for INCF and I am very intrigued by it since it aligns directly with my research interests, so I wanted to keep my options open and ask you about this project as well and how I could go about contributing for it! Do let me know!

I sent you a message in the Orthogonal Lab Slack.

Please join the OpenWorm Slack, join the #devoworm channel, and look over our latest meeting video and thread for more information. Thanks!


@b.alicea I loved the theme of the project. I’m pursuing my bachelors in computer and information science. I have experience in both PyTorch and TensorFlow as I was a contributor in Kaggle working projects ranging from NLP to Forecasting. I’ve taken Biology as minor in associate degree.

How can I start now?

1 Like

Here are some links for people interested in GSoC project 22.1 (GRNs):

A discussion of the GRN project (first 15-20 minutes)

DevoWorm ML:

Media and Public Lectures: Media and Public Lectures

GNN architectures as mazes

GNN for novice mathematicians

DevoZoo (a source fo training data)

DevoLearn paper (Table 1 features another source of data)


Hey, I am Harini, a second year CS undergraduate. I am interested to work on this project.
I have worked with GNNs before as a part of one of my projects. I am also well versed with Pytorch/tensorflow and I think this project is a perfect match. Could we get in touch?

1 Like

Please join the OpenWorm Slack, which I see you already have. Be sure to also join the #devoworm channel and look over the pinned materials for more information on the project.


I would like to join too.
Since I’m planning to do Research at the intersection of AI and Biology.
I’m also very inspired by the words of Demis Hassabis, “Just as mathematics turned out to be the right description language for physics, AI may turn out to play a similar role for biology”. So working on this project would give me a rewarding experience.

1 Like

Here is the invite: OpenWorm Foundation on LaunchPass

1 Like