The 4th Bio-OS AI Open Source Competition of Guangzhou National Laboratory

Competition Introduction

As global science and technology develop rapidly, life science research has entered the era of big data, big platforms, and major discoveries. Big data and artificial intelligence (AI), as important drivers for advancing scientific research, play an increasingly prominent supporting role in areas such as drug and vaccine development, digital organs, multi - omics population cohort studies, and epidemic prevention and control. Breakthrough technological progress and talent reserves in open - source communities are of great significance for fulfilling the strategic mission of Guangzhou National Laboratory.

The Biomedical Big Data Operating System (Bio - OS), independently developed by the Greater Bay Area Bioinformatics Center of Guangzhou National Laboratory, integrates cutting - edge information technology (IT) and artificial intelligence technologies. It serves as a key technological foundation for supporting data - intensive scientific research and AI for Science research. Building on the success of the first two national open - source competitions, the 3rd Bio - OS AI Open - Source Competition is specially held to further support the laboratory's major strategic research directions and fully activate the intellectual resources of the open - source and open community. Focusing on the intersection of AI and biomedicine and AI - driven scientific research innovation, the competition is committed to building a world - class interdisciplinary innovation platform, gathering global wisdom to overcome major technical challenges, accelerating scientific research progress, and promoting technology transfer.

Contest information

Optimization of Secretory Protein Prokaryotic Expression Based on Deep Learning

Aimed at the AI - driven biomedical field, this initiative promotes research through competitions. By leveraging advanced AI technologies, it facilitates the resolution of scientific research challenges in the biomedical domain and enhances China's AI R & D capabilities in biomedicine. It adopts the dry - wet combinationcompetition mode of “AI + experiments” and verifies through wet experiments to ensure the authenticity of competition results.

Protein expression is the process of expressing proteins by means of gene recombination technology. By using systems such as bacteria, yeast, and animal cells, heterologous proteins are expressed to study their structures and functions. There are systemic differences between prokaryotic systems and eukaryotic cells (e.g., different codons). When prokaryotic systems are used to express eukaryotic genes, problems like low expression efficiency and low expression levels tend to occur.

Currently, expressing secretory proteins in prokaryotic systems still poses challenges. It is essential to ensure the solubility, secretion efficiency, and activity of the expressed proteins. However, there is a lack of comprehensive technical guidance on selecting suitable hosts, signal peptides, and molecular tags. In this competition, participants are required to use public datasets to build protein expression algorithms to guide the implementation of secretory protein expression in prokaryotic systems. In the preliminary round, participants need to submit the algorithm for evaluation on the test dataset. In the final round, given a specific secretory protein sequence, they must submit an algorithm - guided vector construction plan. The organizer will conduct experimental validations and assess the expression performance based on the aforementioned indicators.