I am trying to independently analyze raw nanoparticle tracking analysis (NTA) data, but my diffusion coefficients calculated directly from x and y pixel values only moderately correlate with reported values and are consistently off by ~2x.

I attempted subtracting total video average x and y drift (i.e. pixels/frame) and per frame average x and y drift. I attempted fitting mean squared displacement vs time data by just averaging squared displacement, least squares regression with and without nonzero y-intercept, and least squares power law fit (even though anomalous diffusion should not be considered here) all with no better correlation. In fact, even for particle tracks of length 2 frames i.e. a single MSD and time interval, my values still only moderately correlate, depicted below. I have no idea what other variables could account for this variability other than particle intensity, but intensities are rarely used or reported and I have found no significant correlations.

enter image description here

  • 1
    $\begingroup$ Would Cross Validated be a better home for this question? $\endgroup$ – Qmechanic Jan 26 at 16:27
  • $\begingroup$ There are no doubt variables that are un-modeled in your equation. In addition to your log-log plot giving a slope of 2 when it should be 1, the logarithmic scale on both axes means that the errors about the least squares line are MUCH bigger than the plot implies. You need to go back to the raw data, think about ANYTHING that could cause variability in the data, and find a way to include those variables in the regression. $\endgroup$ – David White Jan 26 at 16:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.