[Hend Yamani]

Can anyone please provide me with a reference that use calibration curve other than the original calibration curve of the analyte to calculate LOD?
For example; if the original calibration curve from (1-50 µg/mL), another calibration curve from (0.1-1 µg/mL) is constructed to calculate LOD.

[Hunter2]

I don't know really what you mean. I guess LOD = Limit of detection. And the range is 1-50 µg/l then its not possible to detect below 1 µg/l because of the performance of the analytical method, machine, chemicals, etc. So its not possible to get  reasonable results with a other curve what is constructed below. 0.1-1 µg/l

[Arkcon]

Generally, LOD is defined as 3 times the noise for a representative background interval.  If you want to use a different calibration curve other than the original, I'm guessing you mean to perform a new assay.  Then, you have to re-develop it according to what your noise is.