At iGEM Toronto, our project aims to use generative machine learning to write new, functional DNA sequences that form the template of novel organisms - specifically, new, custom bacteria and phages with controllable phenotypes.
To made headway towards this complex task, we are starting with the simplest proof-of-concept: generating novel plasmid parts.
We are employing a two-pronged approach to make new plasmid parts while creating a positive social impact.
1. Generative AI & Lab Automation
Utilizing AI models, we are generating novel plasmid sequences. These are designed in silico and validated through a comprehensive in silico and in vitro validation pipeline. We are also developing hardware tools to automate this process.
2. Groundwork for Social Impact
Through human practices and entrepreneurship, we are honing in on antibiotic resistance as our first project application and primary commercialization effort. Through these endeavors, we are setting forth a path for Plasmid.AI to positively impact communities globally.
In silico generation and validation
Novel tokenizer and Mamba2 DNA sequence model generates whole plasmids using plasmid dataset. These plasmids are then extensivley validated in silico.
In vitro wet lab validation
The best plasmid parts in silico are put to the test experimentally in the wet lab, where they are assessed for functionality.
IoT tools for high-throughput and lab automation
Modular Internet-of-Things tools, such as colony counters, fridge management systems, and Lean Six Sigma workflows are developed to increase the throughput of the Plasmid.AI pipeline.
Our ultimate goal is refine our proof-of-concept into a platform that can generate, validate, and test entire bacterial and phage genomes, each with fine-tuned functionality that can be used in a myriad of applications; in particular, to solve antibiotic resistance.