SeqTrainer - creating datasets from SynBioHub to aid researchers in creating machine learning models

In the realm of synthetic biology, there is a significant source of data in the form of models, genetic sequences, regulatory networks, and experimental results. Databases such as SynBioHub store this data represented by the Synthetic Biology Open Language (SBOL); using machine learning models, this data can be used by researchers to predict promoter activity, analyze protein-DNA interactions, and more. However, researchers must preprocess SBOL data before training their models, which can be time-consuming. Thus, this project aims to develop a Python package to automate the creation of datasets from data in SynBioHub. In doing so, we will first upload data from previous research to SynBioHub, create the package that returns data from SynBioHub in a format suitable for machine learning models, test the package on the data that was uploaded, explore the performance of a Graph Neural Network (GNN) on the data, and thoroughly document the package.