Preprint, posted 2026 · FakeYeast Consortium

Engineering FakeYeast for tunable fakeGFP output

A synthetic promoter ladder that turns a single fluorescent reporter into a graded, predictable dial across more than a decade of dynamic range, with no antibiotics and no plasmid loss over 200 generations.

Mara Castellanos, Dev Okafor, Lin Wei, Priya Nair, and J. Castellanos Department of Fabricated Genetics, FakeYeast Institute

Explore the figures Data and code Cite

Open data Preprint · not peer reviewed doi 10.0000/fakeyeast.2026.001

0 dynamic range, off to full
0 R squared, Hill fit
0 promoters in the ladder
0 generations, no plasmid loss
Abstract

Fluorescent reporters are usually all or nothing. We rebuilt the fakeGFP expression cassette around a ladder of eight synthetic promoters, each defined by a calibrated ribosome landing pad, so output tracks promoter strength along a smooth Hill curve rather than jumping between states.

The result is a reporter the lab can set to a target brightness the way you set a dimmer. Output is reproducible across colonies, stable for more than 200 generations without selection, and predictable enough that a strength value chosen on this page lands within a few percent of the measured cells. Every strain, construct, and number on this page is fabricated to demonstrate a self hosted paper companion.

Figure 1 · Live 3D

The FakeYeast regulatory interactome

Every gene the promoter ladder touches and how they regulate one another, more than seven hundred genes across five functional modules. Drag to orbit, hover a gene to trace its connections. The fakeGFP cassette sits at the bright core.

0 genes 0 interactions
    Drag to orbit · hover to trace
    Figure 1. The reconstructed FakeYeast regulatory network, nodes colored by functional module, node size scaled by out degree. Rendered live in the browser from the published edge list.
    Figure 2 · Interactive

    fakeGFP output follows a tunable Hill curve

    Each point is a measured strain. Drag the slider to pick a promoter strength and read the predicted output off the fitted curve, the same model the lab uses to design a new strain before building it.

    50 strength 62 AU predicted
    Figure 2. Measured fakeGFP output (points) against promoter strength, with a Hill fit (n = 2.2, K = 45). Shaded band is the 95 percent prediction interval.
    Methods

    How it was built

    Promoter ladder

    Eight cores from a 12,000 member synthetic library were chosen to tile the strength axis evenly on a log scale, each paired with a calibrated ribosome landing pad.

    Measurement

    Output was read by flow cytometry on three biological replicates per strain, normalized to a constitutive marker, with at least 20,000 cells gated per replicate.

    Model

    A two parameter Hill function was fit by least squares. Prediction intervals come from a residual bootstrap with 2,000 resamples.

    Stability

    Strains were passaged for 200 generations without selection. Loss of the cassette stayed below the detection limit at every checkpoint.

    Data and code

    Everything to reproduce the figures

    The strain table, raw cytometry, and the fitting notebook are bundled with this companion. Links are illustrative in the demo.

    Strain Promoter Strength fakeGFP, AU
    strains.csv cytometry_raw.zip hill_fit.ipynb protocol.pdf
    Team

    Who did the work

    Cite this work

    Reference

    Castellanos, M., Okafor, D., Wei, L., Nair, P., and Castellanos, J. (2026). Engineering FakeYeast for tunable fakeGFP output. FakeYeast Consortium preprint. doi 10.0000/fakeyeast.2026.001.