Fast Multi-Task SCCA Learning with Feature Selection for Multi-Modal Brain Imaging Genetics
Lei Du,Kefei Liu,Xiaohui Yao,Shannon L. Risacher,Junwei Han,Lei Guo,Andrew J. Saykin,Li Shen,Michael Weiner,Paul Aisen,Ronald Petersen,Clifford R. Jack,William Jagust,John Q. Trojanowki,Arthur W. Toga,Laurel Beckett,Robert C. Green,John Morris,Enchi Liu,Tom Montine,Anthony Gamst,Ronald G. Thomas,Michael Donohue,Sarah Walter,Devon Gessert,Tamie Sather,Danielle Harvey,John Kornak,Anders Dale,Matthew Bernstein,Joel Felmlee,Nick Fox,Paul Thompson,Norbert Schuff,Gene Alexander,Charles DeCarli,Dan Bandy,Robert A. Koeppe,Norm Foster,Eric M. Reiman,Kewei Chen,Chet Mathis,Nigel J. Cairns,Lisa Taylor-Reinwald,Les Shaw,Virginia M. Y. Lee,Magdalena Korecka,Karen Crawford,Scott Neu,Tatiana M. Foroud,Steven Potkin,Zaven Kachaturian,Richard Frank,Peter J. Snyder,Susan Molchan,Jeffrey Kaye,Joseph Quinn,Betty Lind,Sara Dolen,Lon S. Schneider,Sonia Pawluczyk,Bryan M. Spann,James Brewer,Helen Vanderswag,Judith L. Heidebrink,Joanne L. Lord,Kris Johnson,Rachelle S. Doody,Javier Villanueva-Meyer,Munir Chowdhury,Yaakov Stern,Lawrence S. Honig,Karen L. Bell,Beau Ances,Maria Carroll,Sue Leon,Mark A. Mintun,Stacy Schneider,Daniel Marson,Randall Griffith,David Clark,Hillel Grossman,Effie Mitsis,Aliza Romirowsky,Leyla deToledo-Morrell,Raj C. Shah,Ranjan Duara,Daniel Varon,Peggy Roberts,Marilyn Albert,Chiadi Onyike,Stephanie Kielb,Henry Rusinek,Mony J. de Leon,Lidia Glodzik,P. Murali Doraiswamy,Jeffrey R. Petrella,R. Edward Coleman,Steven E. Arnold,Jason H. Karlawish,David Wolk,Charles D. Smith,Greg Jicha,Peter Hardy,Oscar L. Lopez,MaryAnn Oakley,Donna M. Simpson,Anton P. Porsteinsson,Bonnie S. Goldstein,Kim Martin,Kelly M. Makino,M. Saleem Ismail,Connie Brand,Ruth A. Mulnard,Gaby Thai,Catherine Mc-Adams-Ortiz,Ramon Diaz-Arrastia,Kristen Martin-Cook,Michael DeVous,Allan I. Levey,James J. Lah,Janet S. Cellar,Jeffrey M. Burns,Heather S. Anderson,Russell H. Swerdlow,Liana Apostolova,Po H. Lu,George Bartzokis,Daniel H. S. Silverman,Neill R. Graff-Radford,Francine Parfitt,Heather Johnson,Martin Farlow,Scott Herring,Ann M. Hake,Christopher H. van Dyck,Richard E. Carson,Martha G. MacAvoy,Howard Chertkow,Howard Bergman,Chris Hosein,Sandra Black,Bojana Stefanovic,Curtis Caldwell,Ging-Yuek Robin Hsiung,Howard Feldman,Michele Assaly,Andrew Kertesz,John Rogers,Dick Trost,Charles Bernick,Donna Munic,Diana Kerwin,Marek-Marsel Mesulam,Kristina Lipowski,Chuang-Kuo Wu,Nancy Johnson,Carl Sadowsky,Walter Martinez,Teresa Villena,Raymond Scott Turner,Kathleen Johnson,Brigid Reynolds,Reisa A. Sperling,Keith A. Johnson,Gad Marshall,Meghan Frey,Allyson Rosen,Jared Tinklenberg,Marwan Sabbagh,Christine Belden,Sandra Jacobson,Neil Kowall,Ronald Killiany,Andrew E. Budson,Alexander Norbash,Patricia Lynn Johnson,Thomas O. Obisesan,Saba Wolday,Salome K. Bwayo,Alan Lerner,Leon Hudson,Paula Ogrocki,Evan Fletcher,Owen Carmichael,John Olichney,Smita Kittur,Michael Borrie,T-Y Lee,Rob Bartha,Sterling Johnson,Sanjay Asthana,Cynthia M. Carlsson,Adrian Preda,Dana Nguyen,Pierre Tariot,Adam Fleisher,Stephanie Reeder,Vernice Bates,Horacio Capote,Michelle Rainka,Barry A. Hendin,Douglas W. Scharre,Maria Kataki,Earl A. Zimmerman,Dzintra Celmins,Alice D. Brown,Godfrey D. Pearlson,Karen Blank,Karen Anderson,Robert B. Santulli,Eben S. Schwartz,Kaycee M. Sink,Jeff D. Williamson,Pradeep Garg,Franklin Watkins,Brian R. Ott,Henry Querfurth,Geoffrey Tremont,Stephen Salloway,Paul Malloy,Stephen Correia,Howard J. Rosen,Bruce L. Miller,Jacobo Mintzer,Crystal Flynn Longmire,Kenneth Spicer
DOI: https://doi.org/10.1109/bibm.2018.8621298
2018-01-01
Abstract:Brain imaging genetics studies the genetic basis of brain structures and functions via integrating both genotypic data such as single nucleotide polymorphism (SNP) and imaging quantitative traits (QTs). In this area, both multi-task learning (MTL) and sparse canonical correlation analysis (SCCA) methods are widely used since they are superior to those independent and pairwise univariate analyses. MTL methods generally incorporate a few QTs and are not designed for feature selection from a large number of QTs; while existing SCCA methods typically employ only one modality of QTs to study its association with SNPs. Both MTL and SCCA encounter computational challenges as the number of SNPs increases. In this paper, combining the merits of MTL and SCCA, we propose a novel multi-task SCCA (MTSCCA) learning framework to identify bi-multivariate associations between SNPs and multi-modal imaging QTs. MTSCCA could make use of the complementary information carried by different imaging modalities. Using the G(2,1)-norm regularization, MTSCCA treats all SNPs in the same group together to enforce sparsity at the group level. The l(2,1)-norm penalty is used to jointly select features across multiple tasks for SNPs, and across multiple modalities for QTs. A fast optimization algorithm is proposed using the grouping information of SNPs. Compared with conventional SCCA methods, MTSCCA obtains improved performance regarding both correlation coefficients and canonical weights patterns. In addition, our method runs very fast and is easy-to-implement, and thus could provide a powerful tool for genome-wide brain-wide imaging genetic studies.