BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Turkey BEGIN:DAYLIGHT DTSTART:20380119T061407 TZOFFSETFROM:+0300 TZOFFSETTO:+0300 RRULE:FREQ=YEARLY;BYDAY=3TU;BYMONTH=1 TZNAME:+03 END:DAYLIGHT BEGIN:STANDARD DTSTART:20160907T000000 TZOFFSETFROM:+0300 TZOFFSETTO:+0300 RRULE:FREQ=YEARLY;BYDAY=1WE;BYMONTH=9 TZNAME:+03 END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20191016T083702Z UID:ED0F4E17-8CBD-489B-8FF7-48E2FC17FEEB DTSTART;TZID=Turkey:20191011T133000 DTEND;TZID=Turkey:20191011T153000 DESCRIPTION:11 October 2019 (13:40): IEEE AP/MTT/EMC/ED Turkey Seminar Seri es (S.53)\n\nSpeaker: Asst. Prof. Ercüment Çiçek\, Bilkent University\n \nTopic: "SPADIS: An Algorithm for Selecting Predictive and Diverse SNPs i n Genome-wide Association Studies"\n\nLocation: Middle East Technical Univ ersity\, Ankara\, Turkey\n\nAbstract: Complex traits often cannot be expla ined by individual variants. Therefore\, the efficient selection of multip le loci that explain the phenotype is critical for understanding the genet ic basis of these traits. Selecting multiple loci is a computationally cha llenging problem that grows exponentially with the number of genomic varia nts. Many methods tackle this problem by focusing on coding regions to red uce the complexity of the problem. However\, these approaches ignore the n on-coding regions and introduce literature bias. As one alternative\, regu larized regression methods have been used\; however\, they do not allow th e incorporation of background biological knowledge and suffer from long ex ecution times. Currently\, there is only one machine learning method in th e literature\, which aims to select a large set of loci efficiently by inc orporating biological background information - SConES. SConES selects a se t of features guided by a SNP-SNP network and favors the selection of SNPs that are connected on the network. We argue that while connectedness assu mption is frequently used for functionally related features\, it leads to the selection of redundant features when the goal is to explain a complex phenotype. In the current study\, we hypothesize that selecting features o n an SNP-SNP network that are diverse in term of location would correspond to incorporating complementary terms and thus\, would help to explain the phenotype better. We present SPADIS that implements this novel idea by ma ximizing a submodular set function with a greedy algorithm that ensures a constant factor approximation to the optimal solution. We compare SPADIS t o the state-of-the-art method SConES on a dataset of Arabidopsis Thaliana genotype and continuous flowering time phenotypes. We show that (i) SPADIS has better average phenotype prediction performance in 15 out of 17 pheno types when the same number of SNPs are selected and provides consistent an d statistically significant improvements in regression performance on aver age across multiple networks and settings\, (ii) it identifies more candid ate genes\, and (iii) runs much faster compared to other methods. We also perform rigorous simulation experiments and compare SPADIS with off the sh elf regression-based feature selection methods and show that SPADIS outper forms its counterparts.\n\nBio: Ercument Cicek earned his BS (2007) and MS (2009) degrees in Computer Science and Engineering from Sabanci Universit y. He received his Ph.D. degree in Computer Science from Case Western Rese rve University in 2013. During his Ph.D.\, he visited Cold Spring Harbor L aboratory to work on gene discovery algorithms for Autism Spectrum Disorde r in 2012. After graduation\, he worked as a Lane Fellow in Computational Biology at Carnegie Mellon University till 2015. Since then\, he is an Ass t. Prof. in the Computer Engineering Department of Bilkent University and is an adjunct faculty member in Computational Biology Department of Carneg ie Mellon University. His research is mainly focused on designing machine learning algorithms for analysis of large-scale biological data. He is the recipient of Simons Foundation Autism Research Initiative (SFARI) Explore r Award\, SFARI Pilot Award\, TUBITAK Career Award\, TUBA-GEBIP Award and Parlar Foundation Research Incentive Award.\n\nSpeaker(s): Asst. Prof. Erc üment Çiçek\, \n\nAnkara\, Ankara\, Türkiye LOCATION:Ankara\, Ankara\, Türkiye ORGANIZER:ozergul@metu.edu.tr SEQUENCE:0 SUMMARY:IEEE AP/MTT/EMC/ED TURKEY CHAPTER SEMINAR SERIES -- SEMINAR 53 URL;VALUE=URI:https://events.vtools.ieee.org/m/207440 X-ALT-DESC:Description: <br /><p><strong>11 October 2019 (13:40): &nbsp\;IE EE AP/MTT/EMC/ED Turkey Seminar Series (S.53)</strong></p>\n<p>Speaker: As st. Prof. Erc&uuml\;ment &Ccedil\;i&ccedil\;ek\, Bilkent University</p>\n< p>Topic: "SPADIS: An Algorithm for Selecting Predictive and Diverse SNPs i n Genome-wide Association Studies"</p>\n<p>Location:&nbsp\;Middle East Tec hnical University\, Ankara\, Turkey</p>\n<p>Abstract: Complex traits often cannot be explained by individual variants. Therefore\, the efficient sel ection of multiple loci that explain the phenotype is critical for underst anding the genetic basis of these traits. Selecting multiple loci is a com putationally challenging problem that grows exponentially with the number of genomic variants. Many methods tackle this problem by focusing on codin g regions to reduce the complexity of the problem. However\, these approac hes ignore the non-coding regions and introduce literature bias. As one al ternative\, regularized regression methods have been used\; however\, they do not allow the incorporation of background biological knowledge and suf fer from long execution times. Currently\, there is only one machine learn ing method in the literature\, which aims to select a large set of loci ef ficiently by incorporating biological background information - SConES. SCo nES selects a set of features guided by a SNP-SNP network and favors the s election of SNPs that are connected on the network. We argue that while co nnectedness assumption is frequently used for functionally related feature s\, it leads to the selection of redundant features when the goal is to ex plain a complex phenotype. In the current study\, we hypothesize that sele cting features on an SNP-SNP network that are diverse in term of location would correspond to incorporating complementary terms and thus\, would hel p to explain the phenotype better. We present SPADIS that implements this novel idea by maximizing a submodular set function with a greedy algorithm that ensures a constant factor approximation to the optimal solution. We compare SPADIS to the state-of-the-art method SConES on a dataset of Arabi dopsis Thaliana genotype and continuous flowering time phenotypes. We show that (i) SPADIS has better average phenotype prediction performance in 15 out of 17 phenotypes when the same number of SNPs are selected and provid es consistent and statistically significant improvements in regression per formance on average across multiple networks and settings\, (ii) it identi fies more candidate genes\, and (iii) runs much faster compared to other m ethods. We also perform rigorous simulation experiments and compare SPADIS with off the shelf regression-based feature selection methods and show th at SPADIS outperforms its counterparts.</p>\n<p>Bio: Ercument Cicek earned his BS (2007) and MS (2009) degrees in Computer Science and Engineering f rom Sabanci University. He received his Ph.D. degree in Computer Science f rom Case Western Reserve University in 2013. During his Ph.D.\, he visited Cold Spring Harbor Laboratory to work on gene discovery algorithms for Au tism Spectrum Disorder in 2012. After graduation\, he worked as a Lane Fel low in Computational Biology at Carnegie Mellon University till 2015. Sinc e then\, he is an Asst. Prof. in the Computer Engineering Department of Bi lkent University and is an adjunct faculty member in Computational Biology Department of Carnegie Mellon University. His research is mainly focused on designing machine learning algorithms for analysis of large-scale biolo gical data. He is the recipient of Simons Foundation Autism Research Initi ative (SFARI) Explorer Award\, SFARI Pilot Award\, TUBITAK Career Award\, TUBA-GEBIP Award and Parlar Foundation Research Incentive Award.</p> END:VEVENT END:VCALENDAR