I have been running western blots with multiple biological replicates (separate protein isolations from cells treated the same way) each with multiple technical replicates (different lanes potentially across different gels with protein from the same isolation). I understand that technical replicates are mainly meant to show that one's technique is not too variable, as opposed to biological replicates more showing cellular changes. Using just the means of these technical replicates to create a final distribution seems to throw away information about each one's variability. Is inverse-variance weighting an acceptable or even preferred method to combine all the parameters or should the final distribution only be derived from the means of the technical replicates.
Combine each technical replicates for each biological replicate, then combine each biological replicate into a group (i.e. treated group or baseline group).
Remember technical replicates are there for measuring variability resultant from pipetting, whilst biological replicates for biological variation in expression.
- Badges
- Science method