Question: getIndices identifies non existing cells in a gate
0
gravatar for helena_t
2.8 years ago by
helena_t0
helena_t0 wrote:
Hi,

I have issues with the getIndices() function from the flowWorkspace package. It identifies much more cells as "TRUE" for a gate than it should. Here's my code:

wspath<-"~/Desktop/workspc.wsp"
ws<-openWorkspace(wspath)
G <- parseWorkspace(ws,name=1)
nodelist<-getNodes(G[[1]])
    
num_f5<-grep(pattern = paste("/F5$"),nodelist)
f5<-nodelist[num_f5]
summary(getIndices(G[[1]],f5))​

I have no clue why getIndices() doesn't identify the right cells in the gate I'm interested in, it works perfectly for the gates on top of the gate "F5" in my gating strategy. 

Thanks in advance,

Helena

 

 

 

 

 

 

flowworkspace getindices • 499 views
ADD COMMENTlink modified 2.8 years ago by Jiang, Mike1.2k • written 2.8 years ago by helena_t0
Answer: getIndices identifies non existing cells in a gate
0
gravatar for Jiang, Mike
2.8 years ago by
Jiang, Mike1.2k
(Private Address)
Jiang, Mike1.2k wrote:

First of all, If you just want to get the cells within that gate, you can simply do

getData(G[[1]], f5)

To check if the parsing went well, do

getPopStats(G[[1]])

which will give you the comparison between the flowJo counts and flowWorkspace counts as the validity checks. If they are significantly different (at some gates), then there is parsing problem and you will need to provide more info about the flowjo workspace.

ADD COMMENTlink written 2.8 years ago by Jiang, Mike1.2k
Answer: getIndices identifies non existing cells in a gate
0
gravatar for helena_t
2.8 years ago by
helena_t0
helena_t0 wrote:

The problem is exactly the same with getData(): I have much more F5 cells than expected.

I ran getPopStats and the flowCore counts are clearly very different from the flowJo counts. The flowJo gate that flowWorkspace doesn't manage to read properly is an ellipse. I tried replacing it by a polygon gate. The flowCore counts are closer to the flowJo counts now, but there is still a significant difference.

Is there any argument I could give to getIndices() which would force the function to take the cells that are really in the flowJo gate, and not the cells that flowCore imagines are in the gate?

Thank you in advance,

ADD COMMENTlink written 2.8 years ago by helena_t0

If the gates are not parsed correctly, we then need to fix the parsing first. Can you make sure you have the latest version of flowWorkspace package (i.e. 3.20.4)? If so,  Can you share your example workspace and FCS for me to trouble shoot? (You can send them to wjiang2@fhcrc.org)

ADD REPLYlink written 2.8 years ago by Jiang, Mike1.2k
Answer: getIndices identifies non existing cells in a gate
0
gravatar for helena_t
2.8 years ago by
helena_t0
helena_t0 wrote:

I downloaded the latest version of flowWorkspace (3.20.4), but unfortunately it didn't help. I just sent the FCS file and workspace that FlowWorkspace doesn't parse properly to the email address you indicated.

Thank you for your help

ADD COMMENTlink written 2.8 years ago by helena_t0
Answer: getIndices identifies non existing cells in a gate
0
gravatar for Jiang, Mike
2.8 years ago by
Jiang, Mike1.2k
(Private Address)
Jiang, Mike1.2k wrote:

The fix is in flowWorkspace 3.20.5. Bioconductor should build within 1 day or two.

see the correct `F5` gate here https://github.com/RGLab/flowWorkspace/issues/142#issuecomment-280752179

ADD COMMENTlink written 2.8 years ago by Jiang, Mike1.2k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 307 users visited in the last hour