Electrical and Computer Engineering
menu MENU

CSE Seminar

Using deep-learning methods to predict cancer patient survival

Lana GarmireAssociate Professor of Computational Medicine and BioinformaticsUniversity of Michigan Medical School
WHERE:
3725 Beyster Building
SHARE:

Abstract

Genomics data generally have larger feature sizes than its sample sizes, posing challenges for deep-learning application in this field. In this talk, I will elaborate how we get around the curse of small population size, and apply deep-learning creatively to predict disease prognosis. We have developed a tool called Cox-nnet that uses gene expression data to predict patients survival via neural network. We further developed another integration tool called DeepProg, which uses multiple types of genomics data to predict patients survival via autoencoders and machine learning methods. This method is significantly better than the state of the art method called Similarity Network Fusion (SNF) for molecular subtype identification. We demonstrate the utility of these methods on tens of thousands of cancer samples in the cancer genome atlas (TCGA), and reveal the surprising coherent signatures in all 32 types of cancers. The work builds the foundation for actionable models to extend patient survival time.

Biography

Dr. Garmire is an awardee of US Presidential Early Career Scientisits and Engineers in 2019, the highest honor bestowed to the most outstanding early career scientisits and engineers in the United States. Before joining University of Michigan DCMB department, she rapidly rose to tenure (Dec. 2012 to Jun. 2017) at University of Hawaii Cancer Center, and has become a nationally and internationally recognizable translational bioinformatics scientist leading a multidisciplinary team of computational and experimental human genomics. Garmire has won numerous competitive federal grant awards as the PI, including NIH/NIGMS P20 COBRE (2014-2018, Project Leader), NIH/BD2K K01 award (2014-2021), three concurrent NIH R01 grant awards from NICHD (2016-2021), NLM (2016-2021), and NLM (2019-2022).

Dr. Garmire obtained the MA degree in Statistics (2005) and PhD degree in Comparative Biochemistry (Computational Biology focus, 2007), both from UC-Berkeley. She then did her postdoctoral training (2008-2011) under the joint mentorship of Prof. Shankar Subramaniam in the Bioengineering Dept. and Prof. Christopher Glass in the Department of Cellular and Molecular Medicine, UC-San Diego. In 2012, she won the first bioinformatics NIH SBIR grant for Asuragen Inc (a spin-off of Ambion, the RNA company), and then resumed the tenure-track faculty position since September 2012. Dr. Garmire collaborates with a variety of top researchers nationally and internationally. She has published over 50 papers in high quality journals including Cell and Nature. She has mentored over 40 MD fellows, postdocs, graduate students and undergraduates of various academic backgrounds, in Biology, Mathematics, Phyiscs, (bio)Statistics, Bioengineering, Computer Science and Electrical Engineering. She has served on various NIH study sections. She is an Associate Editor of BMC Bioinformatics and Guest Editor of PLoS Computational Biology.

In her spare time, Dr. Garmire runs around her two young children and an academic husband. She does not have time for hobby except catching up with sleep deprivation.

Organizer

Stephen Reger(734) 764-9401

Faculty Host

Mosharaf ChowdhuryAssistant ProfessorEECS - CSE Division