RO-index

Quantifying fluorescent protein sensor cytGRX-roGFP2 redox state with a reduced–oxidized (RO) index.

Example image: https://nph.onlinelibrary.wiley.com/cms/asset/43d9b2d6-30c5-46de-b688-b91628468678/nph71392-fig-0004-m.jpg

Procedure

1. Load image

1.1 Read confocal microscopy CZI file.

1.2 Extract fluorescence channels:


2. Root region segmentation

For each channel C∈{E405,E488}C \in \{E_{405}, E_{488}\}:

2.1 Normalize intensity values of CC to the range [0,255][0, 255].

2.2 Apply a median filter.

2.3 Compute an Otsu threshold TcT_c.

2.4 Generate binary mask McM_c such that

Mc(x,y)={1,C(x,y)>Tc0,otherwiseM_c(x,y) = \begin{cases} 1, & C(x,y) > T_c \\ 0, & \text{otherwise} \end{cases}

2.5 Retain the masked image

C′=C⋅McC' = C \cdot M_c

3. Protein presence mask

Mprotein=M405∨M488M_{\text{protein}} = M_{405} \lor M_{488}

where ∨\lor denotes the logical OR operator.


4. Saturated pixel removal

For each processed channel C′∈{E405′,E488′}C' \in \{E_{405}', E_{488}'\}:

4.1 Extract all non-zero pixels.

4.2 Compute percentile thresholds

Thigh=P99(C′)T_{\text{high}} = P_{99}(C')

4.3 Ties handling

To avoid too many tied data, only when the ratio of unique intensity values exceeds 0.3, set:

C′(x,y)=0ifC′(x,y)>ThighC'(x,y) = 0 \quad \text{if} \quad C'(x,y) > T_{\text{high}}

5. Redox index calculation

5.1 Square both channels

E405′′=(E405′)2E_{405}'' = (E_{405}')^2 E488′′=(E488′⋅max⁡(E405′)max⁡(E488′))2E_{488}'' = \left( E_{488}' \cdot \frac{\max(E_{405}')}{\max(E_{488}')} \right)^2

5.2 Compute the redox index (RO)

RO=E405′′−E488′′E405′′+E488′′RO = \frac{E_{405}'' - E_{488}''}{E_{405}'' + E_{488}''}

5.3 Replace undefined values with zero.

5.4 Restrict analysis to valid protein regions

RO=RO⋅MproteinRO = RO \cdot M_{\text{protein}}

6. Reduced–Oxidized (RO) index classification

−1≤RO≤1-1 \leq RO \leq 1