2020 NIH iDASH Secure Genome Analysis Competition and Workshop (December 07, 2020, virtual workshop)
Call for Participation Despite the impact of the pandemic, the 7th iDASH Secure Genome Analysis Competition and Workshop is now calling for participation from the academia and the industry to showcase state-of-the-art privacy technologies for protecting real-world biomedical data analysis. In the past 6 years, the iDASH competition has been serving as a bridge between the privacy/security research and biomedical research, challenging the security community to come up with the best solutions that can offer practical supports for privacy-preserving biomedical computing. It has been widely considered to be a benchmark for evaluating data privacy technologies, particularly when they are applied to biomedical data analysis, and a key source for the biomedical and genomics researchers to seek usable solutions for protecting their data and computing tasks. This year’s competition is characterized by 3 tracks as described below. Competition Tasks Track 1: Secure multi-label Tumor classification using Homomorphic Encryption The competitors are required to develop homomorphic encryption (HE) based method for classifying encrypted genetic variant data from tumor samples of unknown type and origin into multiple labels. Track 2: Privacy-preserving clustering of single-cell transcriptomics data in SGX The competitors are expected to implement a trained deep learning model for disease prediction under the protection of SGX, Intel’s trusted execution environment, so the model can work on encrypted genomic data uploaded by the user. Track 3: Differentially private federated learning for the cancer prediction model The competitors are tasked to train a machine learning model on gene expression data for breast tumors, with all the data secretly shared across multiple servers. Timeline 1. Competition start (August 16, 2020) 2. Solution due (October 31, 2020) 3. Winner announcement (December 1, 2020) 4. Workshop day (December 7, 2020) Evaluation The outcomes of the competition will be evaluated by interdisciplinary teams at Indiana University, UC San Diego, and UT Health, based upon the performance of a solution and its privacy guarantee. Organization General Chairs: Arif Harmanci (UT Health), Miran Kim (UNIST), Xiaoqian Jiang (UT Health) Organization Committee: XiaoFeng Wang (IU), Haixu Tang (IU), Xiaoqian Jiang (UT Health), Miran Kim (UNIST), Arif Harmanci (UT Health), Tsung-Ting Kuo (UCSD) and Lucila Ohno-Machado (UCSD) Contact Track 1 (UT Health): Arif Harmanci ([email protected]), Miran Kim ([email protected]<mailto:[email protected]>), Xiaoqiang Jiang ([email protected]<mailto:[email protected]>) Track 2 & 3 (IU): Haixu Tang ([email protected]), XiaoFeng Wang ([email protected]) Best, 2020 iDASH Privacy & Security Workshop organizers
_______________________________________________ Om-announce mailing list [email protected] http://mailman.openmath.org/cgi-bin/mailman/listinfo/om-announce
