CGD-GPT: accelerating medical cannabis research, product development, and clinical practice

Background

My name's Elijah, PhD molecular/cell/dev biology and biotech founder. I built a lab in my garage in 2017 to develop CRISPR engineered cannabis plants, then ended up pivoting and running that business as a diagnostic lab. Before we sunset our services last year, I had launched Cannabis Genome DAO (CGD) in 2021. The goal of this new non-profit entity is to incentivize equitable data sharing between cultivators, researchers, clinicians, and consumers. We're doing that by building a community-governed open-source data marketplace leveraging decentralized ledger technologies and privacy preserving computation provided through use of Ocean Protocol.

https://linkedin.com/in/elijahspinahttps://cannabisgeno.me/

Goal

Our primary goal is to build a tool for scientific researchers and cultivators capable of broadly accelerating the development of new cannabis based therapeutics by simplifying the UX/UI for powerful data analysis models.

Specifically, for this hackathon the CGD team wants to build a chat interface and custom LLM agent trained on our curated dataset. The agent should be capable of utilizing a variety of pre-trained computational biology models and external data to answer complex research questions.

As a chat interface based agent this tool will also be capable of assisting clinicians and consumers in understanding and applying medical cannabis research. A secondary goal would be to incorporate knowledge and data sources sufficient to assist with complex supply chain and ecological research such the tool could assist with efforts to reduce the environmental impact of medical cannabis production.

Starting Point

To prepare, we've been working to build out our training dataset and brainstorming research questions we can use for validation and testing. I’ve also started diving into learning LangChain and exploring a list of pre-trained computational biology models released since I last dove deep into the topic during my phd research.

https://docs.langchain.com/docs/

Team Needs

We currently have a team of four with expert level experience in molecular biology, genomics, cannabis, chemistry, and blockchain; with intermediate experience in data science, computational biology, and predictive modeling.

We’re primarily looking for additional team members who know how to work with data and fine-tune custom agents. Experience in data science, bioinformatics, neural networks (CNNs, GNNs, etc), predictive models using omics data, LLMs, LangChain, or building custom front-ends would be a major plus.

Regardless of previous experience, if you have an interest in the work we’re doing and a desire to contribute, please do feel free to comment below and apply.

Thanks for reading and good luck to everyone participating!