labmatetools / 5.3 BLOT DENSITOMETRY ● IMAGE STAYS ON THIS COMPUTER · NOTHING UPLOADED ALL TOOLS
FIG. 5.3 — BANDS → BOXES → RELATIVE

Blot Densitometry

Band intensities from the raw pixel values — with the image, every box and the arithmetic saved together, so the numbers can be reopened and checked.

IMAGE → LANES → PLACE ROWS → CHECK → NORMALISE → EXPORT

01 — BLOT IMAGETHE ORIGINAL, UNADJUSTED

OR TRY:

02 — LANESLEFT TO RIGHT

Paste a column from Excel or your loading table. Leave out the marker lane — only lanes you measure. Lane names can also come from Protein Loading.

03 — BAND ROWSONE ROW PER PROTEIN

PLACE ROW, then click the middle of the first band and the middle of the last band: a box goes on every lane, evenly spaced. Drag any box to fine-tune it, or select it and use the arrow keys (Shift = 10 px). The dashed box next to each band is its background.

04 — MEASURE

BANDS ARE

As the values are stored, not as you have the display set. Most chemiluminescence exports are dark bands on white; fluorescence scans are usually light on dark.

BACKGROUND

Of the equal-size box next to each band. Median ignores a speck or a scratch in the box; NONE reports raw integrated density.

IMAGE & BOXES

Open a blot image, or try the sample blot.
DISPLAY ONLY · · ·
    RESULTSCLICK A ROW TO SELECT ITS BOX
    WHAT THIS TOOL DOES NOT DO

    Find bands for you · correct over-exposure · compare blots imaged separately · run statistics · tell you whether the experiment worked. It measures the pixels you box, the same way each time.

    The project file (.blot.json) holds the original image, every box, the settings and the algorithm version; open it here to see the same numbers again. The CSV has one line per band with every intermediate value. The ImageJ macro draws the same rectangles in Fiji so you can check RawIntDen yourself.

    REFERENCE · BLOT DENSITOMETRY

    The arithmetic

    QuantityFormula
    Raw integrated densityΣ pixel values in the band box (ImageJ RawIntDen)
    Backgroundmean (or median) of an equal-size box on empty membrane next to the band
    Corrected · light bands on darkΣ − background × area
    Corrected · dark bands on lightbackground × area − Σ
    Normalisedtarget corrected ÷ loading control corrected, same lane
    Relativenormalised ÷ normalised in the reference lane
    Close to backgroundcorrected < 3 × background SD × √area — flagged WEAK

    What gets flagged

    CheckRuleWhy it matters
    Saturatedany band pixel at 0 (dark bands) or at the type maximum — 255, 65535 (light bands)the band’s true value is higher than measured
    Outsidea box reaches past the image edgefewer pixels than the other lanes
    Negativecorrected ≤ 0 — reported as is, never set to zerobox or background in the wrong place, or wrong polarity
    Background overlapa background box touches any band boxbackground too high, band too low
    Loading control spreadhighest ÷ lowest control > 1.5normalisation can hide a loading or transfer problem

    Files it opens

    FormatValues used
    TIFFraw values: 8 / 16 / 32-bit grey (incl. float); uncompressed, LZW, PackBits, Deflate; first page of a stack. White-is-zero files are flipped to match the display.
    PNGraw values: 1–16-bit grey, read by this page (no browser colour or gamma correction)
    RGB TIFF / PNGone channel if R = G = B; otherwise (R + G + B) ÷ 3, as ImageJ measures RGB — flagged
    JPEG · WebP · BMP8-bit as decoded by the browser — JPEG and WebP flagged as lossy
    Not opened.scn, .1sc, .gel and other imager formats; tiled or BigTIFF — export a TIFF first

    Worked example

    Example. A 3-pixel box over a dark band holds 110, 120 and 130 (Σ = 360); its background box averages 200. Corrected = 200 × 3 − 360 = 240. If the β-actin band in the same lane is 400, the normalised value is 240 ÷ 400 = 0.60; if the reference lane’s normalised value is 0.75, this lane is 0.80 relative.

    LAST UPDATED · HOW VALUES ARE CHECKED · ABOUT LABMATE.TOOLS

    QUESTIONS · BLOT DENSITOMETRY

    Frequently asked

    Is my blot image uploaded?

    No. The image is opened and measured in your browser and never uploaded. Only your settings, and the box positions for the last image, are kept on this device.

    Which image should I quantify?

    The original from your imager, exported as a 16-bit TIFF if it offers one, without contrast or brightness changes. 8-bit and JPEG exports are often already adjusted or compressed, which changes the numbers.

    How is the band value calculated?

    Integrated density (the sum of pixel values in the box, ImageJ’s RawIntDen) minus the background value times the box area. For dark bands on a light background the sign is flipped: background × area − sum.

    Why must all boxes in a row be the same size?

    Integrated density grows with the area of the box. Equal boxes for one protein make the lanes comparable; the tool keeps every box in a row at the same width and height.

    What does a saturated pixel mean?

    A pixel at the detector’s limit (0 for dark bands, 65535 or 255 for light ones) has lost information, so the band is underestimated. Re-image with a shorter exposure; the tool flags every band with such pixels.

    How do I normalise to a loading control?

    Add a row for the loading control (β-actin, GAPDH, tubulin or a total-protein stain) and set NORMALISE TO on the target row. Each lane’s target value is divided by the control in the same lane, then expressed relative to the reference lane.

    Can I check the numbers in ImageJ or Fiji?

    Yes. Download the ImageJ macro, open the same unchanged image in Fiji and run it: it draws the same rectangles and measures them, and RawIntDen should match raw_int_den in the CSV.

    Can I reopen the analysis later?

    Yes. The project file holds the original image, every box, the settings and the algorithm version. Open it here and the numbers are recomputed and compared with the saved ones.