FCS Background Noise Fitting Tool

This tool estimates the background noise of a flow cytometer channel by fitting a Gaussian distribution to the low‑intensity portion of the signal. It then reports the fitted mean (μ), standard deviation (σ), and a default threshold (LOD) defined as μ + 5σ. The plot shows:

• All Data — histogram of all events on a single, common bin grid,
• Filtered for fit — the subset of those same bins used for the fit,
• Fit — the Gaussian curve estimated from the filtered region,
• Threshold — a vertical line at LOD (μ + 5σ).

The “Filtered for fit” is an actual subset of “All Data” using the same bin edges, ensuring μ and σ are directly comparable to what you see in the full histogram.

1. Upload your .fcs file in FCS File.
2. Select a Channel (e.g., FSC, SSC, or a fluorescence channel).
3. (Optional) Set an Upper cutoff (a.u.) to exclude very bright events; it’s clamped between 1 and the channel’s max value.
4. Adjust Bins Noise Factor for a smoother or finer histogram.
5. (Optional) Toggle Force Fit Through Zero if your baseline is known to be 0.
6. Inspect the plot and the Fitting Results (μ, σ, LOD).

Channel Selection

Choose which detector/channel to analyze. Each channel can have a different range and background.

Upper cutoff (a.u.)

Removes events with values ≥ the cutoff before histogramming (default 100,000). Useful to prevent high‑intensity tails from biasing the background fit. The input is automatically constrained to [1, max(channel data)].

Bins Noise Factor

Controls histogram resolution by setting an approximate bin width. Larger numbers → fewer, wider bins (smoother but less detailed); smaller numbers → more, narrower bins (more detail but noisier).

Force Fit Through Zero

Fits only the right side of the histogram and constrains μ = 0. Use when your instrument baseline is physically anchored at zero.

Threshold multiplier (σ)

Select how many standard deviations from the mean you would like to put the threshold. In our experience five has been a safe option. If you dont mind including background noise you can lower this number.

Optional: Diameter Channel Selection

Select the channel that contains your calibrated diameter data. This data is used to interpolate the threshold from arbitrary units to standard units.

• If the blue Filtered for fit region tracks the gray All Data near the baseline and the dashed black fit overlays it well, your background is well‑described by a Gaussian.
• The red dotted threshold is the default LOD (μ + 5σ). Points to the right of this line are increasingly unlikely to be background under the Gaussian model.
• If the fit is visibly poor (e.g., skewed tails, multiple peaks), try increasing the upper cutoff, adjusting the bin factor, or toggling “Force Fit Through Zero.”

• Pick a channel and gate settings representative of background (e.g., blank or negative controls).
• Start with the default upper cutoff; raise it if too little data is included or lower it if too much data is included.
• Use a moderate bin factor. If the histogram looks jagged, increase it; if it hides structure, decrease it.
• “Force through zero” is best when you analyse the height parameter because the background distribution of this parameter has an offset.
• Large files: analysis runs entirely in your browser; performance depends on your device.

• Threshold = μ + 5σ is a recommendation based on our findings; your assay may require a different multiplier.
• Your data is processed locally in the browser; files are not uploaded to a server by this tool.