Meeting_Minutes_Lowe_Sander

First Meeting @ 3/5/2012 5:00 pm Attendees: Chris Sander, Giovanni, Rileen, Scott Lowe, Vishal, Thomas Next meeting is on April 6th 2012 at 9:30 AM. Work on coassociations along with Giovanni's work on Mutual Exclusivity will be presented then. Meeting April 6th: Minutes and summary of talks
 * 1) Scott talked about various projects and "Biological Engine" that Lowe lab has for testing candidate genes via pooled screen or otherwise
 * 2) Chris talked about some of the tools and the Mutual Exclusivity project done by Giovanni
 * 3) The co-association (Deletion and Amplification) idea was expressed by Scott. Vishal, Rileen and Giovanni to meet on 3/6/2012 to discuss on POA for this
 * Nicholas Gauthier**: (Grad std)
 * Cell selevtive labeling of proteine sin co-culture
 * C-TAP: Motivation: Biomarker discovery: Hyp: if we can detect proteins that are secreted by tumor cells, they might be better biomarkers
 * 1) Mass spectrometer.
 * 2) How do we know which prots come thru cancer cells
 * SILAC: stable isotope labeling of amino acids in cell culture
 * Control (state A) Preturbed (State b, heavy)
 * Mix them and then the MS can identify proteins.
 * Conclusions**
 * 1) Cell selective
 * 2) continuous labeling
 * 3) co-cultures in vitro
 * 4) tissue or cell specific labeling in vivo?
 * 5) canonical amino acids
 * 6) non biased
 * 7) high throughput
 * 8) Quantitative

Interrelations of Genomic events in Cancer Mutual Exclusivity: MEmo: mutual exclusivity Model Feature 1: Altered genes frequently than expected 2) Genes in same pathway 3) Genes altered in mutually exclusive (TCGA nature 2011) CNA and mutations. (GISTIC/RAE) Background network: Pathway commons. This helps us filter the tests. Leaves out unknown alterations. Working on that. Pairwise events tested (1 tumor type vs others: Always samples are analysed 1 at a time) DO analysis for multiple cancers Permutations done on: Conclusions cBIO portal. Amplicons, focal amplicons. 8p loss, equals something like a p53 mutation. Biology of 8p loss. Driver of that loss. Gene set to test of co-occurence Don't test for single genes. Test is for amplicons. GISTIC Algorithm: Frequently, first filter correlation bet mRna and copy number. Drawback: Extracts, focal regions. If you have a whole army band, it doesnt account for that.
 * Giovanni Ciriello**
 * alterations per gene
 * alterations per sample
 * include other tupes of alterations, Mrna methylations
 * test co-occurence
 * unknown associations?
 * So far: test all possible pairwise associations
 * NExt: Algorithmically infer functional relationship

Anil Korkut Perturbation Cell biology: Network models of drug respnce in cancer cells

pathway commons: BioPax and Paxtools Off target effect NFKbeta paper Modeling of off target effects of shRNA's

( From Rileen: Actually, the Sander group is supposed to talk about some projects next time, not just association or exclusivity - but if we have some promising preliminary results by then, we can certainly present them)