I have got Attune NxT flow cytometry data in the latest version of flowJo (PC) that I wish to use FlowClean, downsample and t-SNE on.
All three are listed in my plug-ins and when selected none of them work correctly. For flowclean and downsample a dialogue box pops up for me to enter variables of interest, once OK is clicked flowclean or downsample logos appear (see attached screen shot for example) next to my gates but, following 'calculation' data does not appear for me to utilise....e.g. there is no data showing in the statistics or count column ..... has anybody else had this problem and overcome it? Thanks Adam
Hi Adam, I have not tried yet to use the FlowJo implementation of FlowClean, but both downsampling and tSNE analysis works just fine. The Downsampling should actually give you a "*.pop" gate that you can use for further analysis.
Did you follow all the instructions on the FlowJo homepage for installing plugins? You also have to make sure that you install the R environment (for some plugins) and point FlowJo to the correct path: http://docs.flowjo.com/d2/plugins/installing-plugins/
Should you run a tSNE analysis, be aware that this will take a while, on a typical deskop computer 50.000 cells will probably take 5-10 minutes of computing time.
Dear Florian, Apologies for the late response and many thanks for your answer above. I have managed to get downsample to work but I do need to follow your suggestions regarding R...Since the number of cells can be specified by the user, I assume this will have a major impact on the results e.g. t-SNE based on 3,000 vs 50,000 cells. How have you decided on the threshold/number of cells that are most appropriate? best wishes, Adam
Hi Adam, the overall pattern of the tSNE plot that you get should be same regardless of whether you use 3000 or 50.000 cells - however, you might miss populations that are present only at low frequency.
In the end, running more cells if of course beneficial, but requires more computation time (on a typical desktop computer for 50.000 cells will take anything between 5-15 minutes). There is an upper limit, at which you would run into the so-called crowding problem (see original publications on tSNE and also a couple of reviews), but that should not be problem if you stay below 100.000 cells.
By the way, if you want to compare different samples, make sure that you concatenate them together beforehand! Have fun with your analyis :)
Thanks Florian....I have recently been successful in running Flow Clean, Downsampling and tSNE and indeed enjoying the features....particularly the heat map in the latest flowJo version...thanks for the advice. I may come back when I get to the stage of comparing longitudinal samples...
Hi Adam,
I have the same problem you mentioned in the OP. How did you solve it? At times, flowjo says tSNE (calculating), but it goes away if I move the mouse. Thank you for any insights.
Sam
Hi Adam,
I have the same problem that you mentioned above when I try to perform the downsampling.. only once it worked, when I had one sample on the workspace.. could it be an events number trouble? I work with tumor.. I have soo many events, but I do downsampling on a gated population..
Thanks!
Sabri
Hi, Adam
I have problems with several plugins with FlowJo which need to run R, for example SPADE. Depending on the size of your samples, the Downsample and tSNE take very long time to calculate.
I would be grateful if anyone run SPADE with FlowJo PC version to share their experience.
I have to use FCS Express 6 for SPADE analysis.
Best wishes
Yanping
- Badges
- Science topic