Skip to main content
AboutGuideContact
Sign in / Register

Flow Cytometry Compute

Most flow cytometry work in Conspecta runs in your browser. Two operations, dimensionality reduction and clustering, get heavy enough on real datasets that you can offload them to the cloud. Those server runs are the only flow operations that consume compute usage.

Local vs Server

When you open the Dimensionality Reduction or Clustering dialog box, the bottom of the panel shows a Compute Location toggle:

  • Local runs the algorithm in your browser. It's free and works well on small to medium event counts.
  • Server sends the job to a cloud worker. Use this for large datasets that would lock up your tab. Server jobs consume compute usage.

The toggle defaults to Server when the dataset is large enough that the local run would be slow. You can flip it manually at any time.

The dialog box shows the estimated usage next to the Server option before you click Run, so there are no surprises.

What Consumes Compute Usage

Only these flow operations consume usage, and only when Server is selected:

  • Dimensionality reduction with UMAP, t-SNE, or PCA
  • Clustering with K-means, DBSCAN, FlowSOM, Phenograph, or Density

That's the entire list. Every other flow feature is free.

What's Always Free

Every other flow operation runs in your browser for free:

  • Gating, gate hierarchy editing, and population math
  • Compensation matrix calculation and the compensation wizard
  • Plotting (dot plots, contour plots, histograms, density plots)
  • Statistics, summary tables, MFI comparisons
  • File concatenation, batch statistics
  • Specialized analyses like cell-cycle, proliferation, and kinetics
  • All exports (FCS, Excel, PDF, plot images)

Running a Server Job

  1. Open the Dimensionality Reduction or Clustering dialog box from the analysis toolbar
  2. Pick the algorithm and parameters
  3. Confirm Server is selected in the Compute Location panel
  4. Review the estimated usage
  5. Click Run

The dialog box shows progress while the job is queued, then while it's running. You can leave the dialog box open and keep working on other things; the result is fetched back into the analysis when it's done.

Tracking and Cancelling Jobs

Every server job appears in your team's Compute Jobs list (Settings → Usage → Compute Jobs). Each row shows:

  • The job type (Flow Computation)
  • Who started it and when
  • Status (queued, running, completed, failed, cancelled)
  • Usage

You can also export the full list as CSV from that panel.

Next Steps

Plans and billingImage analysis compute