- Cellarity and CZI are committed to driving innovation in single-cell methods development through supporting the Open Problems science initiative and united in the belief that advances in analytical approaches are critical for ensuring growth of the single-cell field and impacting human health
- This partnership provides a continued commitment to support open science initiatives with the Open Problems in Single-Cell Analysis consortium, including hosting a Kaggle competition for NeurIPS 2023 challenging scientists to efficiently model simulated drug discovery experiments
Cellarity, a life sciences company founded by Flagship Pioneering to transform the way medicines are created, and the Chan Zuckerberg Initiative (CZI) today announced a partnership to drive innovation in machine learning (ML) algorithms for single-cell analysis via support of the Open Problems in Single-Cell Analysis initiative. Starting with a collaboration in 2021 to host a multimodal data integration competition, CZI and Cellarity’s partnership has grown to support efforts that include the creation of unique benchmarking datasets to power competitions drawing thousands of participants and the creation and support of community-driven, open-source algorithm benchmarks hosted at Open Problems.
“Single-cell biology is at the forefront of biomedical research and data science. Now more than ever, we recognize the importance of benchmarks and challenges to help bring data science and cutting-edge computational approaches to single-cell biology,” said Jonah Cool, Ph.D., Senior Program Officer of the Single-Cell Biology program at the Chan Zuckerberg Initiative. “We are excited to again support Cellarity and Open Problems in Single Cell Analysis as we unravel the mysteries of the cell, moving beyond descriptive biology and into mechanistic understandings.”
Fabrice Chouraqui, Pharm.D., CEO of Cellarity and CEO-Partner at Flagship Pioneering, said, “At Cellarity, we are constantly pushing the boundaries of machine learning applications for single-cell analysis. We are proud to partner with CZI and Open Problems in Single-Cell Analysis to unlock new applications that enable us to create medicines more efficiently and bring innovations to patients faster. We are excited to build on our track record of success hosting machine learning competitions in 2021 and 2022 to, this year, challenge machine learning experts around the world to model simulated drug discovery experiments more efficiently.”
“There are more than 1,500 algorithms developed for single-cell data, and understanding the deep complexity of cells captured by single-cell technologies requires robustly evaluating the performance of these methods,” says Diogo Camacho, Ph.D., Vice President of Computational Biology at Cellarity. “It’s exciting to see Cellarity scientists partner with academic leaders around the world to develop fair and open benchmarks.”
For the first competition in 2021 featuring multimodal data integration, Cellarity collaborated with the Chan Zuckerberg Biohub San Francisco, Yale University, and Helmholtz Münich to generate the largest realistic benchmarking dataset currently available for multimodal single-cell data, defining three distinct tasks for multimodal single-cell data integration and metrics to evaluate them. The consortium hosted a second competition in 2022 that was powered by a new benchmarking dataset, a reformulation of the task metrics, and an extension to temporal dynamics modeling. The 2022 competition attracted more than 1,600 competitors on Kaggle and is, to the best of our knowledge, the largest single-cell competition ever conducted.
This year’s Kaggle competition will tackle a new challenge in single-cell data science and introduce another novel benchmark dataset to tackle problems in modeling cellular response to chemical perturbation. The objective is to develop methods that can generalize to unseen perturbations and cell types to enable scientists to overcome the practical and economic limitations of single-cell perturbation studies. The goal of this competition is to leverage advances in representation learning to unlock new capabilities bridging data science, machine learning, and computational biology.
Entries to the competition will be accepted until November 30, 2023. For more information, visit the competition page on Kaggle.
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!
