plz. Clarify the meaning of
manupulation in your terms - "should not be done at all"
and refinement - " there are certain allowed changes to figures etc one can do which varies according to individual journal"
as far as I know" ".
But, science & research will be meaningless if one can't publish truth.
Are you asking about the number of replicates for each experiment and deciding statistical significance?
Or are you asking typically how many submission and review iterations it takes to get a manuscript accepted in a scientific journal? There is no set answer to that question. It varies dramatically from submission to submission and journal to journal and is heavily dependent on the quality of your initial submission.
Peer review exists to provide constructive criticism on your manuscript so that even if it is deemed unworthy of publication on your first attempt you learn what is neccessary to improve the quality of your next submission.
i guess refinements as in improving picture quality(this applies to the entire picture and not to specific regions!) is ok to a certain extent, however, manipulations or refinements which might result in misinterpretation of data and misrepresentation of results is definitely not allowed.....the best idea would be to take maximal care when scanning your blots or clicking images, so that one does not have to make any changes later on..