Predicting the location of proteins within cells reveals a wealth of biological information important for researchers to develop future scientific discoveries related to drug development and treatment of diseases such as epilepsy. help. That’s because proteins are the body’s “workhorse products” and are primarily responsible for most cellular functions.
Recently, Dong Xu, Distinguished Curatorial Professor in the Department of Electrical Engineering and Computer Science, University of Missouri, and colleagues updated the protein localization prediction model MULocDeep to provide more targeted predictions, including animal and human specific models. made available. and plants. The model was originally created by Xu and his MU researcher and biochemistry professor Jay Thelen ten years before him to study proteins in mitochondria.
“Many biological discoveries need to be verified by experiments, but researchers don’t have to spend time and money doing thousands of experiments to get there,” Xu said. “A more targeted approach saves time. Our tools can help researchers design more targeted experiments to advance their research more effectively, so they can get to discovery faster.” By making it accessible, it provides a valuable resource for researchers.”
The model harnesses the power of artificial intelligence through machine learning techniques (training a computer to make predictions using existing data) to generate protein irregularities known as “false localizations”. It can assist researchers studying the position, or mechanisms associated with where a protein goes. To a place different from the original. This abnormality is often associated with diseases such as metabolic disorders, cancer, and neurological disorders.
“Some diseases are caused by mislocalization, where proteins either fail to reach their targets or reach their targets inefficiently, thus failing to function as expected,” Xu said. rice field.
Another application of the team’s predictive model is to aid drug design by targeting misplaced proteins and moving them to the correct location, Xu said.
This research is currently supported by the National Science Foundation. In the future, Xu hopes to get additional funding to improve the model’s accuracy and develop more features.
“We continue to refine our model to determine whether protein mutations can cause mislocalization, whether proteins are distributed in multiple cellular compartments, or whether signal peptides can localize more precisely. We want to determine how it can help us predict the future,” Xu said. “Although we do not provide solutions for drug development or cures for various diseases per se, our tools may help other companies develop medical solutions. It’s like, different people play different roles, and together we can achieve a lot for everyone.”
Xu is currently working with colleagues to develop a free online course for high school and college students based on the biology and bioinformatics concepts used in the model, which will be available later this year. is going to be
Xu and others also point out conflicts of interest. An online version of MULocDeep is available for academic users, but a standalone version is also available commercially for a license fee.
“MULocDeep Web Service for Protein Localization Prediction and Visualization at the Subcellular and Subcellular Organelle Level” was published in the journal Nucleic Acids Research. Co-authors are Yuexu Jiang, Lei Jiang, Chopparapu Sai Akhil, Duolin Wang, Ziyang Zhang, and Weinan Zhang from MU.
