Year: 2025
Type: preprint
Abstract: Gene circuit documentation is key in synthetic biology. After conceiving and designing a model, scientists need an effective tool to map, and later, implement such circuits in the wet-lab. Most software applications that tackle this problem, generate SBOL 2 outputs, even in SBOL 3.1 is the current version. In this context, we present GC-AI as a solution designed to simplify gene circuit conversion. It transforms a circuit from the whiteboard or any other handwritten source into an equivalent file in SBOL 3.1 standard. This is achieved in three stages: circuit scanning, structural part definition, and functional part definition. The first stage is solved by using AIbased image processing: a pretrained model of the PyTorch library: Faster R-CNN with ResNet50-FPN. The second stage links the output prediction of the circuit to user-selected components found in the iGEM parts registry. The third and final stage defines functions and interactions between parts. Thus, GC-AI is a solution that eases and accelerates the design of gene circuits into SBOL 3.1 standard. Also, GC-AI is capable of transforming the SBOL 3.1 output to FASTA and/or GenBank formats.
Institution: Universidad Diego Portales
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: gold
OpenAlex ID: W4413105290