[Colloquium] Seminar Announcement: Using patient networks to integrate genomic data

Ninfa Mayorga ninfa at uchicago.edu
Thu Oct 16 14:43:56 CDT 2014


~Reminder~

Computation Institute Presentation

Speaker: Anna Goldenberg, Assistant Professor, Department of Computer Science, University of Toronto, Canada, Scientist, Genetics & Genome Biology Program, SickKids
Host:  Natalia Maltsev
Date:  October 17, 2014
Time: 10:30 AM - 11:30 AM
Location: University of Chicago, Searle 240A, 5735 S. Ellis Ave.

Using patient networks to integrate genomic data
Combining multiple types of measurements to create a comprehensive view of a given disease or a biological process has become an important computational challenge. In this talk I will introduce Similarity network fusion (SNF) to integrate genomic and other types of data for the same set of subjects or objects. For a given disease, SNF  solves the integration problem by constructing networks of patients for each available data type and then efficiently fusing these into one network that represents the full spectrum of the underlying data. Combining mRNA expression, DNA methylation and microRNA (miRNA) expression data, SNF substantially outperformed single data type analysis and established integrative approaches to identify cancer subtypes in five cancer data sets. Extending Cox regression by incorporating patient-network regularization we showed a substantial improvement in concordance index compared to the regular subtyping analysis on the METABRIC dataset.  We have applied SNF in multiple contexts effectively combining continuous and discrete, clinical and genomic data to identify new disease subtypes and have most recently extended this work to partially complete measurements: for example, SNF is now applicable in scenarios where the measurements were collected for partially overlapping cohorts of patients.  This talk will be largely (but not completely) based on Wang et al, Nature Methods, 2014.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20141016/36355ccc/attachment-0001.htm 


More information about the Colloquium mailing list