Everything. Finding someone to complete the work (and revisions), digging up the data (even if it was organized), repeating experiments, coordinating across multiple people esp when one of them has moved on to a different position. Just very complicated.
I guess one good thing about most of my research being done by undergrads or technicians for the last 10 or so years is that I just expect I'll be finishing the paper myself, so we try to wrap each experiment up completely as we go.
Things I require before someone can 'leave' an experiment:
Replications and statistical analysis
Illustrator file with figure legend
Record of where I can find notes and data for the publication-quality image and the biological repeats.
Excel file with all reagents used.
Methods for the figure.
Comments
Make sure everything is publication-ready now. It'll make it easier to publish if you aren't in the lab in the future.
Replications and statistical analysis
Illustrator file with figure legend
Record of where I can find notes and data for the publication-quality image and the biological repeats.
Excel file with all reagents used.
Methods for the figure.
I generally have to help people a lot when they are starting out, but most undergrads leave my lab competent in making figures.
I also have a figure template, which makes things a little easier.