[flowCAP] flowcap

[Woda, Marcia]

Dear Sirs/Madam

 

Pardon my ignorance, but I have a few questions. Is the challenge ongoing?  If one submits a data set, is there any feedback on computational tools you are using that work well with said dataset.

 

Thank you

 

Marcia Woda

UMMS

Worcester, Ma.

01655

--
You received this message because you are subscribed to the Google Groups "FlowCAP" group.
To post to this group, send email to flo...@googlegroups.com.
To unsubscribe from this group, send email to flowcap+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/flowcap?hl=en.

[Ryan Brinkman]

Marcia,

Yes, the current competitions are ongoing. The first deadline for submission is the end of next month, and we have already received entries, so it will be hard to get new data into this round. On the plus side, we are already planning to have another competition, and feedback would include information on what approaches work well the dataset. For the next competition we hope to have the gold standard that the automated approaches have to match would be related to sort of outcome data (e.g., patient survival), post gating. This would avoid some of the deficiencies in making comparisons to something we know has flaws (manual gating), and where the comparison is setup such that it is impossible for the automated approaches to score "better". We also think that taking the competition one step past gating, we will be getting to the real questions people want answered, which isn't necessarily where to draw the boxes around groups of cells. Of course where to place the gates in the next set of challenges will still be an important part, but not the be-all and end-all.

So, if you have data where you also have some associated outcome information (e.g., these set of samples are from sick patients, these are from healthy; these samples were treated with compound x and different doses), then it would be great to include these in the next round. The goal in this case would be to see how different approaches are for finding important differences across treatments, etc. Perhaps its best we talk about what sorts of data you have and how we could include it next round.

Best,
Ryan

On 2010-05-20, at 1:20 PM, Woda, Marcia wrote:

Dear Sirs/Madam

Pardon my ignorance, but I have a few questions. Is the challenge ongoing? If one submits a data set, is there any feedback on computational tools you are using that work well with said dataset.

Thank you

Marcia Woda
UMMS
Worcester, Ma.
01655

--
You received this message because you are subscribed to the Google Groups "FlowCAP" group.

To post to this group, send email to flo...@googlegroups.com<mailto:flo...@googlegroups.com>.
To unsubscribe from this group, send email to flowcap+u...@googlegroups.com<mailto:flowcap+u...@googlegroups.com>.