Home Career Researchers develop user-friendly software system for optimizing biological systems — ScienceDaily

Researchers develop user-friendly software system for optimizing biological systems — ScienceDaily

43
0

Machine learning is transforming all areas of life science and industry, but is typically limited to a few users and scenarios. A team of researchers from the Max Planck Institute for Terrestrial Microbiology, led by Tobias Erb, has developed METIS, a modular software system for the optimization of biological systems. The research team demonstrates its convenience and versatility with various biological examples.

Although biological systems engineering is indeed indispensable in biotechnology and synthetic biology, today machine learning has become useful in all areas of biology. However, it is clear that the application and improvement of algorithms, computational procedures compiled from lists of instructions, are not easily accessible. They are not only limited by programming skills, but often also by insufficient experimentally labeled data. At the intersection of computational and experimental work, there is a need for efficient approaches to bridge the gap between machine learning algorithms and their application to biological systems.

Now a team from the Max Planck Institute for Terrestrial Microbiology, led by Tobias Erb, has succeeded in democratizing machine learning. In their recent publication in Nature Communications, the team presented their METIS instrument with collaboration partners from the INRAe Institute in Paris. The application is built in such a universal and modular architecture that it does not require computing skills and can be used in different biological systems and with different laboratory equipment. METIS is short for Machine Learning-led Experimental Trials for Improvement of Systems, and is also named after the ancient goddess of wisdom and crafts, Μῆτις, letters. “wise advice”.

Less data is required

Active learning, also known as optimal experimental design, uses machine learning algorithms to interactively suggest the next set of experiments after learning from previous results, a valuable approach for scientists in wet labs, especially when working with limited experimentally labeled data. But one of the main bottlenecks is the experimentally labeled data obtained in the laboratory, which are not always high enough to train machine learning models. “While active learning already reduces the need for experimental data, we went further and explored different machine learning algorithms. What is encouraging is that we have found a model that is even less dependent on the data,” says Amir Pandey, one of the study’s lead authors.

To show the versatility of METIS, the team used it for a variety of applications, including optimizing protein production, genetic constructs, combinatorial engineering of enzyme activity, and complex CO2 a fixed metabolic cycle called CETCH. For the CETCH cycle, they explored a combinatorial space of 1025 conditions from only 1000 experimental conditions and reported the most efficient CO2 fixation cascade described to date.

Optimization of biological systems

In application, the research offers new tools to democratize and advance current efforts in biotechnology, synthetic biology, genetic engineering, and metabolic engineering. “METIS allows researchers to either optimize their already discovered or synthesized biological systems,” says Christoph Diehl, one of the study’s lead authors. “But it’s also a combinatorial aid to understanding complex interactions and hypothesis-driven optimization. And what is probably the most interesting advantage: it can be a very useful system for prototyping new-to-nature systems.”

METIS is a modular tool that works like the Google Colab Python notebooks and can be used through a personal copy of the notebook in a web browser, with no installation, no registration, and no need for local computing power. The materials presented in this work can help users configure METIS for their applications.

Story source:

Materials is provided Max Planck Society. Note: Content can be edited for style and length.

Source link

Previous articleNo Secrets: Loudoun Co. School Board. denies Miyares’ application to bar the public from hearing the lawsuit
Next articleFinancial help offered to local businesses to recruit young people – FE News