Initial installation of Scaffold in R

[Nyaradzo]

Hello,

I am struggling to get started with using scaffold in R. I have been able to load all the packages but when i get to loading my files nothing happens. I get the NULL response in R studio and I cannot choose a dataset or anything on the page in the browser. Can you help?

Nyari

[vjcitn]

I am not a developer of this code but I am commenting as an interested party.

I would agree that the README, while strong on installation concepts, does not really take a new user through the steps required to have a successful first run. You are using Rstudio. Thus when you start run.scaffold(), a browser window to drive the analysis will open, but also, a widget will pop up asking you to pick a file. Implicitly you must pick a file that has .fcs as suffix. In the browser interface, you are asked in the README to select "Run clustering". Given that you have chosen an FCS file, the "Run clustering" panel will have dropdowns with options for file choice, marker choice, and clustering options. Certain defaults are provided. I used an FCS file from flowCore: R RHOME/library/flowCore/extdata/0877408774.B08 ... it must be renamed to have a .fcs suffix to be identified in scaffold. It is taking forever to cluster on the markers I chose, so I think I must have done something wrong. It would be good for the developers to supply a test dataset where "answers" are clear for certain selections of tuning parameters.

[SamGG]

I am developing this neither, but testing Scaffold is in my todo list.

I fully agree that a small set of FCS files would allow one to test the installation and the concept. FCS files usually end with a ".fcs" extension. Nolan lab is mainly working with latest CyTOF instrument whereas flowCore demo files were acquired a decade ago or so.

One must select a FCS file in order to inform the software about the location of the folder that contains the FCS files to process (using shinyFiles library sounds a better option IMHO, https://cran.r-project.org/web/packages/shinyFiles). In order to restart a shiny app, press the red STOP of the R console.

To reproduce/accelerate Vince's test, select FL[1234]-H channels only, ask for 20 clusters instead of 200, set arcsinh to 150 or 500 rather than 5 (150 is for fluorescent cytometry and stated later in the README), and run (I don't know what "samples" means, may be "runs" ie repetitions). Within a few minutes you get text files as results. Unfortunately, I get stuck in constructing a SCAFFoLD map.

HTH

> scaffold.run()

Listening on http://127.0.0.1:4659
NULL
[1] "Performing clara clustering"
[1] "Clustering done"
[1] "Performing clara clustering"
[1] "Clustering done"
[1] "Performing clara clustering"
[1] "Clustering done"
[1] "Markers used for SCAFFoLD: FL1-H, FL2-H, FL3-H, FL4-H"
[1] "Using as reference 0877408774.B08.fcs.clustered.txt"
[1] "Downsampling to 1000 events"
Warning: Error in <-: attempt to set an attribute on NULL
Stack trace (innermost first):
    89: load_attractors_from_gated_data
    88: scaffold:::run_analysis_gated
    81: isolate
    80: renderText [C:\Users\samgg\Documents\R\win-library\3.3\scaffold\shinyGUI/server.R#352]
    79: func
    78: origRenderFunc
    77: output$analysisui_empty
     2: runApp
     1: scaffold.run

Characteristics of demo files.

flowFrame object '0877408774.B08'
with 10000 cells and 8 observables:
     name              desc range minRange maxRange
$P1 FSC-H             FSC-H  1024        0     1023
$P2 SSC-H             SSC-H  1024        0     1023
$P3 FL1-H              <NA>  1024        0     1023
$P4 FL2-H              <NA>  1024        0     1023
$P5 FL3-H              <NA>  1024        0     1023
$P6 FL1-A              <NA>  1024        0     1023
$P7 FL4-H              <NA>  1024        0     1023
$P8  Time Time (51.20 sec.)  1024        0     1023
149 keywords are stored in the 'description' slot

Archive with FCS files and clustering results scaffold-demo.zip

[zinagood]

Hi all,

I'm a PhD student in the nolan lab and have successfully used scaffold once by following instructions. Since academic labs don't usually tend to have full-time staff for tech support and programming, the code is on GitHub for the purpose of being examined and extended. If you'd like to get any test files, our lab usually deposits those with publication. For example, you can download FCS data here: https://web-stanford-edu.laneproxy.stanford.edu/~samusik/Panorama BM 1-10.zip

Hope you find this helpful! And good luck!! Zina

-- Zinaida Good 415-290-8944

On Thu, Mar 16, 2017 at 1:12 PM, SamGG notifications@github.com wrote:

I am developing this neither, but testing Scaffold is in my todo list.

I fully agree that a small set of FCS files would allow one to test the installation and the concept. FCS files usually end with a ".fcs" extension. Nolan lab is mainly working with latest CyTOF instrument whereas flowCore demo files were acquired a decade ago or so.

One must select a FCS file in order to inform the software about the location of the folder that contains the FCS files to process (using shinyFiles library sounds a better option IMHO, https://cran.r-project.org/web/packages/shinyFiles). In order to restart a shiny app, press the red STOP of the R console.

To reproduce/accelerate Vince's test, select FL[1234]-H channels only, ask for 20 clusters instead of 200, set arcsinh to 150 or 500 rather than 5 (150 is for fluorescent cytometry and stated later in the README), and run (I don't know what "samples" means, may be "runs" ie repetitions). Within a few minutes you get text files as results. Unfortunately, I get stuck in constructing a SCAFFoLD map.

HTH

scaffold.run()

Listening on http://127.0.0.1:4659 NULL [1] "Performing clara clustering" [1] "Clustering done" [1] "Performing clara clustering" [1] "Clustering done" [1] "Performing clara clustering" [1] "Clustering done" [1] "Markers used for SCAFFoLD: FL1-H, FL2-H, FL3-H, FL4-H" [1] "Using as reference 0877408774.B08.fcs.clustered.txt" [1] "Downsampling to 1000 events" Warning: Error in <-: attempt to set an attribute on NULL Stack trace (innermost first): 89: load_attractors_from_gated_data 88: scaffold:::run_analysis_gated 81: isolate 80: renderText [C:\Users\samgg\Documents\R\win-library\3.3\scaffold\shinyGUI/server.R#352] 79: func 78: origRenderFunc 77: output$analysisui_empty 2: runApp 1: scaffold.run

Characteristics of demo files.

flowFrame object '0877408774.B08' with 10000 cells and 8 observables: name desc range minRange maxRange $P1 FSC-H FSC-H 1024 0 1023 $P2 SSC-H SSC-H 1024 0 1023 $P3 FL1-H 1024 0 1023 $P4 FL2-H 1024 0 1023 $P5 FL3-H 1024 0 1023 $P6 FL1-A 1024 0 1023 $P7 FL4-H 1024 0 1023 $P8 Time Time (51.20 sec.) 1024 0 1023 149 keywords are stored in the 'description' slot

Archive with FCS files and clustering results scaffold-demo.zip https://github.com/nolanlab/scaffold/files/848907/scaffold-demo.zip

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-287177057, or mute the thread https://github.com/notifications/unsubscribe-auth/AIDe2w6SoE-p2DR0aQbOZe9oD5MOiOtKks5rmZeVgaJpZM4MfPyN .

[Nyaradzo]

Dear all,

Thank you for your help. Sorted now.

Regards

Nyari

From: Zinaida Good [mailto:notifications@github.com] Sent: 17 March 2017 05:16 AM To: nolanlab/scaffold Cc: Nyari Chigorimbo; Author Subject: Re: [nolanlab/scaffold] Initial installation of Scaffold in R (#2)

Hi all,

I'm a PhD student in the nolan lab and have successfully used scaffold once by following instructions. Since academic labs don't usually tend to have full-time staff for tech support and programming, the code is on GitHub for the purpose of being examined and extended. If you'd like to get any test files, our lab usually deposits those with publication. For example, you can download FCS data here: https://web-stanford-edu.laneproxy.stanford.edu/~samusik/Panorama BM 1-10.zip

Hope you find this helpful! And good luck!! Zina

-- Zinaida Good 415-290-8944

On Thu, Mar 16, 2017 at 1:12 PM, SamGG notifications@github.com<mailto:notifications@github.com> wrote:

I am developing this neither, but testing Scaffold is in my todo list.

I fully agree that a small set of FCS files would allow one to test the installation and the concept. FCS files usually end with a ".fcs" extension. Nolan lab is mainly working with latest CyTOF instrument whereas flowCore demo files were acquired a decade ago or so.

One must select a FCS file in order to inform the software about the location of the folder that contains the FCS files to process (using shinyFiles library sounds a better option IMHO, https://cran.r-project.org/web/packages/shinyFiles). In order to restart a shiny app, press the red STOP of the R console.

To reproduce/accelerate Vince's test, select FL[1234]-H channels only, ask for 20 clusters instead of 200, set arcsinh to 150 or 500 rather than 5 (150 is for fluorescent cytometry and stated later in the README), and run (I don't know what "samples" means, may be "runs" ie repetitions). Within a few minutes you get text files as results. Unfortunately, I get stuck in constructing a SCAFFoLD map.

HTH

scaffold.run()

Listening on http://127.0.0.1:4659 NULL [1] "Performing clara clustering" [1] "Clustering done" [1] "Performing clara clustering" [1] "Clustering done" [1] "Performing clara clustering" [1] "Clustering done" [1] "Markers used for SCAFFoLD: FL1-H, FL2-H, FL3-H, FL4-H" [1] "Using as reference 0877408774.B08.fcs.clustered.txt" [1] "Downsampling to 1000 events" Warning: Error in <-: attempt to set an attribute on NULL Stack trace (innermost first): 89: load_attractors_from_gated_data 88: scaffold:::run_analysis_gated 81: isolate 80: renderText [C:\Users\samgg\Documents\R\win-library\3.3\scaffold\shinyGUI/server.R#352] 79: func 78: origRenderFunc 77: output$analysisui_empty 2: runApp 1: scaffold.run

Characteristics of demo files.

flowFrame object '0877408774.B08' with 10000 cells and 8 observables: name desc range minRange maxRange $P1 FSC-H FSC-H 1024 0 1023 $P2 SSC-H SSC-H 1024 0 1023 $P3 FL1-H 1024 0 1023 $P4 FL2-H 1024 0 1023 $P5 FL3-H 1024 0 1023 $P6 FL1-A 1024 0 1023 $P7 FL4-H 1024 0 1023 $P8 Time Time (51.20 sec.) 1024 0 1023 149 keywords are stored in the 'description' slot

Archive with FCS files and clustering results scaffold-demo.zip https://github.com/nolanlab/scaffold/files/848907/scaffold-demo.zip

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-287177057, or mute the thread https://github.com/notifications/unsubscribe-auth/AIDe2w6SoE-p2DR0aQbOZe9oD5MOiOtKks5rmZeVgaJpZM4MfPyN .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/nolanlab/scaffold/issues/2#issuecomment-287255944, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AQo9_OslUKZKuq0ltIRG63msrS2o77f8ks5rmfr6gaJpZM4MfPyN.

[vjcitn]

On Thu, Mar 16, 2017 at 11:16 PM, Zinaida Good notifications@github.com wrote:

Hi all,

I'm a PhD student in the nolan lab and have successfully used scaffold once by following instructions. Since academic labs don't usually tend to have full-time staff for tech support and programming, the code is on GitHub for the purpose of being examined and extended. If you'd like to get any test files, our lab usually deposits those with publication. For example, you can download FCS data here: https://web-stanford-edu.laneproxy.stanford.edu/~samusik/Panorama BM 1-10.zip

Thanks for your note. Is that link resolvable to those outside Stanford? I am getting a login prompt for a Stanford-linked institution.

Hope you find this helpful! And good luck!! Zina

-- Zinaida Good 415-290-8944 <(415)%20290-8944>

On Thu, Mar 16, 2017 at 1:12 PM, SamGG notifications@github.com wrote:

I am developing this neither, but testing Scaffold is in my todo list.

I fully agree that a small set of FCS files would allow one to test the installation and the concept. FCS files usually end with a ".fcs" extension. Nolan lab is mainly working with latest CyTOF instrument whereas flowCore demo files were acquired a decade ago or so.

One must select a FCS file in order to inform the software about the location of the folder that contains the FCS files to process (using shinyFiles library sounds a better option IMHO, https://cran.r-project.org/web/packages/shinyFiles). In order to restart a shiny app, press the red STOP of the R console.

To reproduce/accelerate Vince's test, select FL[1234]-H channels only, ask for 20 clusters instead of 200, set arcsinh to 150 or 500 rather than 5 (150 is for fluorescent cytometry and stated later in the README), and run (I don't know what "samples" means, may be "runs" ie repetitions). Within a few minutes you get text files as results. Unfortunately, I get stuck in constructing a SCAFFoLD map.

HTH

scaffold.run()

Listening on http://127.0.0.1:4659 NULL [1] "Performing clara clustering" [1] "Clustering done" [1] "Performing clara clustering" [1] "Clustering done" [1] "Performing clara clustering" [1] "Clustering done" [1] "Markers used for SCAFFoLD: FL1-H, FL2-H, FL3-H, FL4-H" [1] "Using as reference 0877408774.B08.fcs.clustered.txt" [1] "Downsampling to 1000 events" Warning: Error in <-: attempt to set an attribute on NULL Stack trace (innermost first): 89: load_attractors_from_gated_data 88: scaffold:::run_analysis_gated 81: isolate 80: renderText [C:\Users\samgg\Documents\R\win-library\3.3\scaffold\ shinyGUI/server.R#352] 79: func 78: origRenderFunc 77: output$analysisui_empty 2: runApp 1: scaffold.run

Characteristics of demo files.

flowFrame object '0877408774.B08' with 10000 cells and 8 observables: name desc range minRange maxRange $P1 FSC-H FSC-H 1024 0 1023 $P2 SSC-H SSC-H 1024 0 1023 $P3 FL1-H 1024 0 1023 $P4 FL2-H 1024 0 1023 $P5 FL3-H 1024 0 1023 $P6 FL1-A 1024 0 1023 $P7 FL4-H 1024 0 1023 $P8 Time Time (51.20 sec.) 1024 0 1023 149 keywords are stored in the 'description' slot

Archive with FCS files and clustering results scaffold-demo.zip https://github.com/nolanlab/scaffold/files/848907/scaffold-demo.zip

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-287177057, or mute the thread https://github.com/notifications/unsubscribe-auth/AIDe2w6SoE- p2DR0aQbOZe9oD5MOiOtKks5rmZeVgaJpZM4MfPyN .

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-287255944, or mute the thread https://github.com/notifications/unsubscribe-auth/AEaOwrqFgnXnktTozuBDhllcjug7ez4Hks5rmfr4gaJpZM4MfPyN .

[DMrdjen]

Hey guys, How can I export pdf's of the scaffold maps?

[davemcilwain]

Download SVG crowbar

On Thu, Mar 30, 2017 at 7:16 AM, Dunja Mrdjen notifications@github.com wrote:

Hey guys, How can I export pdf's of the scaffold maps?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-290424226, or mute the thread https://github.com/notifications/unsubscribe-auth/AICG6iJi3Lflsx00YEhIM_xoI5kGDkYyks5rq7ktgaJpZM4MfPyN .

-- David R. McIlwain, PhD Postdoctoral Fellow Garry Nolan Lab Baxter Laboratory for Stem Cell Biology Stanford University Medical Center 269 Campus Drive, CCSR 3235 Stanford, CA 94305-5175 United States

Email: davemcilwain@gmail.com Mobile: +1(650) 646-8957 Web: https://med.stanford.edu/profiles/david-mcilwain

[DMrdjen]

Cool thanks for the tip.

---

Dunja Mrdjen University of Zurich Institute of Experimental Immunology Laboratory of Burkhard Becher Lab Y44 Room J94, Office Y23 K67 Winterthurerstr. 190 CH-8057 Zurich Phone: +41-44-63-53707 / 53718 mrdjen@immunology.uzh.ch mailto:mrdjen@immunology.uzh.ch www.immunology.uzh.ch http://www.immunology.uzh.ch/

On 30 Mar 2017, at 22:19, davemcilwain notifications@github.com wrote:

Download SVG crowbar

On Thu, Mar 30, 2017 at 7:16 AM, Dunja Mrdjen notifications@github.com wrote:

Hey guys, How can I export pdf's of the scaffold maps?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-290424226, or mute the thread https://github.com/notifications/unsubscribe-auth/AICG6iJi3Lflsx00YEhIM_xoI5kGDkYyks5rq7ktgaJpZM4MfPyN .

-- David R. McIlwain, PhD Postdoctoral Fellow Garry Nolan Lab Baxter Laboratory for Stem Cell Biology Stanford University Medical Center 269 Campus Drive, CCSR 3235 Stanford, CA 94305-5175 United States

Email: davemcilwain@gmail.com Mobile: +1(650) 646-8957 Web: https://med.stanford.edu/profiles/david-mcilwain — You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-290532686, or mute the thread https://github.com/notifications/unsubscribe-auth/ATLIvYHr_g46ySFdOLSsIco78yym79vbks5rrA5WgaJpZM4MfPyN.

[DMrdjen]

Is it possible to plot single cells instead of clustered nodes, in points rather than in bubbles?

[zinagood]

Hi Dunja, The single-cell force-directed layout has been imlitemned in Vortex software (Samusik et al, Nature Methods 2016). Yo can include landmark cells there. To my knowledge, there are no plan to implement singe-cell view in Scaffold (it may make Scaffold plots too messy also). Zina

-- Zinaida Good 415-290-8944

On Mon, May 22, 2017 at 1:51 AM, Dunja Mrdjen notifications@github.com wrote:

Is it possible to plot single cells instead of clustered nodes, in points rather than in bubbles?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/nolanlab/scaffold/issues/2#issuecomment-303039328, or mute the thread https://github.com/notifications/unsubscribe-auth/AIDe2yJ9ua9-MEul4yaH9rDNO-ega8BXks5r8UyagaJpZM4MfPyN .

[Nemo2013]

Hi there, This is the first time I am using scaffold. I have got the .Rdata files and am trying to run the scaffold analysis as in the README. I keep getting an error "missing value where TRUE/FALSE needed". I have tried filling in different combinations of things in the drop down menu/tick boxes but still get this error. I have my gated fcs files in a subdirectory "gated"...

This is the print out I get in R studio - tack trace (innermost first): 89: load_attractors_from_gated_data 88: scaffold:::run_analysis_gated 81: isolate 80: renderText [/Library/Frameworks/R.framework/Versions/3.3/Resources/library/scaffold/shinyGUI/server.R#352] 79: func 78: origRenderFunc 77: output$analysisui_empty 2: runApp 1: scaffold.run

Any suggestions where I am missing this TRUE/FALSE value?

Thanks very much. Kirsten

[mvetill]

Hi ! I'm also experiencing problem to lunch scaffold since the installation went well, the select file window popup as well as the browser, but when I choose a FCS file from a fcs file folder nothing happen.

library(scaffold) scaffold.run()

Listening on http://127.0.0.1:5732

here windows popup normally and I choose a FCS file

ERROR: [on_request_read] connection reset by peer

please Help :) Regards!!

[SamGG]

Hi! @mvetill I think you should open a new issue and tell the computer OS, the web browser, the R version and any message that appeared in the console. A Google search points to many causes, but you should try to use Firefox or Chrome if you are not. HTH