As part of its growing engagement in data science and artificial intelligence for mining bioimages, Leica Microsystems is proud to sponsor the upcoming Kaggle competition Human Protein Atlas 2018 Image Challenge.
Participants should develop new methods for analysis of large bioimage datasets. Manually determining the localization of proteins in cells is time-consuming and laborious. Using new approaches of data science and deep learning to analyze an ever greater number of bioimages can lead to faster, more accurate cell protein localization.
The goal is to build a program that automatically classifies confocal microscope image data of proteins in vertebrate cells into about 30 different organelle classes. Those participants achieving the most successful results are eligible to win cash prizes and a NVIDIA GV100 GPU. Additionally, the journal Nature Methods has indicated interest to consider publication of a paper discussing the best performing model.
The data to be used for this competition stems from the Human Protein Atlas. By identifying the cell organelles where a protein is localized, its function and potential interactions with other proteins and biomolecules can be better understood. In the future, the resulting model could help to further clarify cellular mechanisms important to human health. A possible consequence is that better treatments for certain diseases may be discovered.
The Cell Atlas, a part of the Human Protein Atlas, was created by the group of Emma Lundberg at the SciLifeLab, KTH Royal Institute of Technology, in Stockholm, Sweden. Currently, she is a visiting professor at Stanford University through the support of the Chan Zuckerberg Initiative. The Cell Atlas was created, in large part, using data acquired with Leica confocal instruments.
Through sponsoring this competition, Leica Microsystems is further able to contribute to both the extension and improvement of biological knowledge, as well as, help build tools for precisely analyzing the vast amount of data created.
Individuals interested to take part in the competition can find more detailed information at Human Protein Atlas 2018 Image Challenge.