Fast and Scalable Private Genotype Imputation Using Machine Learning and Partially Homomorphic Encryption.
Fast and Scalable Private Genotype Imputation Using Machine Learning and Partially Homomorphic Encryption.
复制标题
DOI：
10.1109/access.2021.3093005
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Maniatakos M
中科院分区：
文献类型：
--
作者：
Sarkar E;Chielle E;Gürsoy G;Mazonka O;Gerstein M;Maniatakos M
基因组测序技术的最新进展为理解人类基因变异和疾病之间的关系提供了前所未有的机会。然而，对一大群个体的整个基因组进行基因分型仍然是成本高昂的。预测缺失基因变异的基因类型的填补方法被广泛使用，特别是在全基因组关联研究中。准确的基因分型需要复杂的统计方法。由于该问题的数据和计算密集型性质，归罪越来越多地被外包，这引发了严重的隐私问题。在这项工作中，我们使用机器学习(ML)和一种标准化的同态加密方案Paillier密码系统来研究快速、可扩展和准确的隐私保护基因归属的解决方案。基于ML的隐私保护推理已经在很大程度上针对单输出多类分类设置中计算繁重的非线性函数进行了优化。然而，每个个体的每个基因组具有大量的多类输出需要针对这一应用进一步的优化和/或近似。在这里，我们探索了使用标准化的同态加密方案，将用于基因归罪的线性模型转换为保护隐私的等价物的有效性。我们的结果表明，我们的隐私保护的基因推算方法的性能相当于最先进的明文解决方案，即使在高达8万个目标的真实世界大型数据集上，也可以获得高达99%的曲线下微区域。
The recent advances in genome sequencing technologies provide unprecedented opportunities to understand the relationship between human genetic variation and diseases. However, genotyping whole genomes from a large cohort of individuals is still cost prohibitive. Imputation methods to predict genotypes of missing genetic variants are widely used, especially for genome-wide association studies. Accurate genotype imputation requires complex statistical methods. Due to the data and computing-intensive nature of the problem, imputation is increasingly outsourced, raising serious privacy concerns. In this work, we investigate solutions for fast, scalable, and accurate privacy-preserving genotype imputation using Machine Learning (ML) and a standardized homomorphic encryption scheme, Paillier cryptosystem. ML-based privacy-preserving inference has been largely optimized for computation-heavy non-linear functions in a single-output multi-class classification setting. However, having a large number of multi-class outputs per genome per individual calls for further optimizations and/or approximations specific to this application. Here we explore the effectiveness of linear models for genotype imputation to convert them to privacy-preserving equivalents using standardized homomorphic encryption schemes. Our results show that performance of our privacy-preserving genotype imputation method is equivalent to the state-of-the-art plaintext solutions, achieving up to 99% micro area under curve score, even on real-world large-scale datasets up to 80,000 targets.
登录
查看更多内容
影响因子：
30.8
作者：
Das, Sayantan;Forer, Lukas;Schoenherr, Sebastian;Sidore, Carlo;Locke, Adam E.;Kwong, Alan;Vrieze, Scott I.;Chew, Emily Y.;Levy, Shawn;McGue, Matt;Schlessinger, David;Stambolian, Dwight;Loh, Po-Ru;Iacono, William G.;Swaroop, Anand;Scott, Laura J.;Cucca, Francesco;Kronenberg, Florian;Boehnke, Michael;Abecasis, Goncalo R.;Fuchsberger, Christian
通讯作者：
Fuchsberger, Christian
影响因子：
8
作者：
Bradley, AP
通讯作者：
Bradley, AP
影响因子：
16.6
作者：
Naveed, Muhammad;Ayday, Erman;Wang, Xiaofeng
通讯作者：
Wang, Xiaofeng
影响因子：
30.8
作者：
Lonsdale, John;Thomas, Jeffrey;Salvatore, Mike;Phillips, Rebecca;Lo, Edmund;Shad, Saboor;Hasz, Richard;Walters, Gary;Garcia, Fernando;Young, Nancy;Foster, Barbara;Moser, Mike;Karasik, Ellen;Gillard, Bryan;Ramsey, Kimberley;Sullivan, Susan;Bridge, Jason;Magazine, Harold;Syron, John;Fleming, Johnelle;Siminoff, Laura;Traino, Heather;Mosavel, Maghboeba;Barker, Laura;Jewell, Scott;Rohrer, Dan;Maxim, Dan;Filkins, Dana;Harbach, Philip;Cortadillo, Eddie;Berghuis, Bree;Turner, Lisa;Hudson, Eric;Feenstra, Kristin;Sobin, Leslie;Robb, James;Branton, Phillip;Korzeniewski, Greg;Shive, Charles;Tabor, David;Qi, Liqun;Groch, Kevin;Nampally, Sreenath;Buia, Steve;Zimmerman, Angela;Smith, Anna;Burges, Robin;Robinson, Karna;Valentino, Kim;Bradbury, Deborah;Cosentino, Mark;Diaz-Mayoral, Norma;Kennedy, Mary;Engel, Theresa;Williams, Penelope;Erickson, Kenyon;Ardlie, Kristin;Winckler, Wendy;Getz, Gad;DeLuca, David;MacArthur, Daniel;Kellis, Manolis;Thomson, Alexander;Young, Taylor;Gelfand, Ellen;Donovan, Molly;Meng, Yan;Grant, George;Mash, Deborah;Marcus, Yvonne;Basile, Margaret;Liu, Jun;Zhu, Jun;Tu, Zhidong;Cox, Nancy J.;Nicolae, Dan L.;Gamazon, Eric R.;Im, Hae Kyung;Konkashbaev, Anuar;Pritchard, Jonathan;Stevens, Matthew;Flutre, Timothee;Wen, Xiaoquan;Dermitzakis, Emmanouil T.;Lappalainen, Tuuli;Guigo, Roderic;Monlong, Jean;Sammeth, Michael;Koller, Daphne;Battle, Alexis;Mostafavi, Sara;McCarthy, Mark;Rivas, Manual;Maller, Julian;Rusyn, Ivan;Nobel, Andrew;Wright, Fred;Shabalin, Andrey;Feolo, Mike;Sharopova, Nataliya;Sturcke, Anne;Paschal, Justin;Anderson, James M.;Wilder, Elizabeth L.;Derr, Leslie K.;Green, Eric D.;Struewing, Jeffery P.;Temple, Gary;Volpi, Simona;Boyer, Joy T.;Thomson, Elizabeth J.;Guyer, Mark S.;Cathy Ng;Abdallah, Assya;Colantuoni, Deborah;Insel, Thomas R.;Koester, Susan E.;Little, A. Roger;Bender, Patrick K.;Lehner, Thomas;Yao, Yin;Compton, Carolyn C.;Vaught, Jimmie B.;Sawyer, Sherilyn;Lockhart, Nicole C.;Demchok, Joanne;Moore, Helen F.
通讯作者：
Moore, Helen F.
影响因子：
30.8
作者：
Howie, Bryan;Fuchsberger, Christian;Stephens, Matthew;Marchini, Jonathan;Abecasis, Goncalo R.
通讯作者：
Abecasis, Goncalo R.