Active Projects

 Theme 1   Scalable ML algorithms for high-dimensional data
 Theme 2   Predictive medicine
 Theme 3   Biomarker discovery, personalized medicine
 Theme 4   Molecular biology, evolutionary biology, cancer biology





Machine learning in the operating room       

about: The goal of this project is to develop a ML system to predict respiratory events from surgery and anesthesia in real time, based on time-series measurements of patient's condition during the surgery
people: Scott Lundberg, Su-In Lee, Jerry Kim, Bala Nair, Shu-Fang Newman
funding: ITHS, eScience



Efficient dimensionality reduction for high-dimensional network estimation            

about: A
For more details, please read Celik et al. presented at the NIPS Workshop on MLCB 2013 and ICML 2014.
people: Safiye Celik, Benjamin Logsdon, Su-In Lee
funding: NSF, AAUW (Safiye's fellowship), NIH



A probabilistic framework for meta-analysis of high-dimensional network inference            

about: 
funding: NSF



Identifying molecular markers for tumor resectability                 

about: 
people: Benjamin Logsdon, Su-In Lee, David Hawkins, Stephanie Battle, Mara Rendi, Charles Drescher
funding: Solid Tumor Translational Research, NIH



Network-based classification of driver vs. passenger mutations       

about: 
people: Patricio Velez, Su-In Lee
funding: NIH



An unsupervised feature learning approach to classify cancer patients                 

about: 
funding: Solid Tumor Translational Research



Identifying expression drivers in cancer            

about: 
people: Benjamin Logsdon, Su-In Lee, Andrew Gentles, Chris Miller, Tony Blau, Pamela Becker
funding: NIH



A big data approach to identify molecular markers for chemotherapy drugs in AML            

about: 
people: Benjamin Logsdon, Su-In Lee, Vivian, Oehler, Chris Miller, Tony Blau, Pamela Becker
funding: NIH, Life Sciences Discovery Fund


and many more (under construction)...



 Identifying network perturbation
 about    
 people  


cancer systems biology:
learning biological networks that can define cancer subtypes and predict treatment outcomes.
Ben Logsdon, Su-In Lee
collaborators: Andrew Gentles (Stanford), David Hawkins (Genome Sciences), Tony Blau (Medicine), Pay Monnat (Pathology).


personalized cancer treatment: building a prediction model to select the best chemotherapy drugs for individual patients.
collaborators: Tony Blau (Medicine), Pam Becker (Medicine).


systems genetics: identifying the complex path from genotype to phenotype.
collaborators: Daphne Koller (Stanford), Sulin Wu (ULCA), Tom Drake (UCLA), Jake Lusis (UCLA)


understanding the effect of hormone therapy on gene networks
Ben Logsdon, Su-In Lee
collaborators: Nick Smith, Bruce Psaty, Astrid Suchy-Dicey (Cardiovascular Health Research Unit)



human-friendly machine learning
Danielle Bragg, James Forgarty, Su-In Lee



retreving information from 3D image data for medical diagnosis
Ezgi Mercan, Linda Shapiro, Su-In Lee.
collaborators: medical investigators at U of Pittsburgh



classification and interest region localization on craniosynostosis skulls
Lynn Yang, Su-In Lee, Linda Shapiro
collaborators: Michael Cunningham, Matthew Speltz, Craig Birgfeld, Indriyati Atmosukarto (Seattle Children's Research Institute).

predictive medicine: applying ML techniques for better prediction of patients' sudden condition changes.
collaborators: Meliha Yetisgen-Yildiz (BHI), Heather Evans (Dept. of Surgery), Emily Fox (Statistics).




genetics of Alzheimer's disease and personalized treatment
Ben Logsdon, Su-In Lee
collaborators: Tom Montine (Neuropathology), Paul Crane (UW Medicine).


identifying driver pathways in cancer
Maxim Grechkin, Su-In Lee
collaborators: Andrew Gentles (Stanford), Tony Blau (Medicine), Ray Monnat (Pathology).



learning interaction between different layers in gene regulation
Patricio Velez, Larry Ruzzo, Su-In Lee
collaborators: Tony Blau (UW Medicine), Ray Monnat (Pathology).



network-based understanding of complex diseases and treatment outcomes
Sonya Alexandrova, Su-In Lee 
collaborators: Nathan Price (ISB).


understanding the genetic and molecular basis of drug sensitivity phenotype
Peter Ney, Su-In Lee
collaborators: Maitreya Dunham (Genome Sciences), Tony Blau (UW Medicine), Ray Monnat (Pathology).


sparse learning of very large graphical models
Safiye Celik, Ben Logsdon, Su-In Lee




structured joint graphical lasso
Karthik Mohan, Mike Chung, Seungyeop Han, Su-In Lee.
collaborators: Daniela Witten (Biostatistics), Maryam Fazel (EE)

sparse learning of Gaussian graphical models
Karthik Mohan, Palma London, Su-In Lee.
collaborators: Daniela Witten (Biostatistics), Maryam Fazel (EE)

and more.
Subpages (2): MGL PamNet