Sunday, June 30, 2013

CROWDSOURCING


 

CROWDSOURCING: A new tool.



In an English tavern.
No, that's not Guiness.  It's her coke.
 
 
Perhaps inspired by  various TV reality shows, the NCI has instituted a program wherein a cancer-related “problem” is defined and “teams” from around the globe invited to work on it.  Then, after the lapse of some time, NCI (or somebody) decides who came up with the best solution and rewards them – hold your breath – with a trip to San Francisco and a chance to publish a peer-reviewed paper in a prestigious medical journal.  It may seem like I'm scoffing, but I'm not:  I think it is a very good idea. 
The projects I just read about are part of a larger program called DREAM:  Dialog for Reverse Engineering, Assessments and Methods.  DREAM engages in “crowdsourcing”, which means taking advantage of expertise and knowledge across the research community to attack specific problems.  NCI-DREAM is a subset  focused on several aspects of breast cancer.  Apparently 52 teams elected to participate.  They were given a standard genome, then invited to: (1) using the genomic data and, I guess, anything else they could drum up, predict  the response of 18 breast cancer “lines” to 31 previously untested drugs, and (2) predict the activity of pairs of compounds on diffuse large B-cell lymphoma (again using a standard cell line.)  The value of this sort of research to individualized cancer treatment is pretty obvious.
A team from Helsinki won sub-challenge (1); a team from Texas . #2.    
There also is another challenge afoot, one to predict breast cancer survival rates.  It uses something called “computational biology”.  Around every corner in a cancer research institution one bumps into a statistician, a computer programmer, or this thing called a computational biologist.  And to think I  used to look down on biologists as hopelessly non-quantitative!  Like, geology uses higher math?
P.S.  I am back from Borrego Springs until after Christmas.
P.P.S.  I have used 88 pictures so far, and I am beginning to forget which.  If you catch me doubling up, please let me know - I have many I want to use.  For those who don't know:  myrlbeck@msn.com



No comments:

No comments:

Post a Comment