The amount of proteins (ug per sample) and also the number of replicates you need to run for a quantitative proteomics experiment depends on different things.
1. What type of analysis will you do? 2D GE with MALDI ToF MS identification, targeted LC-MS or untargeted LC-MS, LC-MS quantification with labeled proteins (SILAC, Di-methylated, 18O, TMT etc) or label free quantification?
2. What type of sample do you have? (Serum, tissue, cell culture etc.) do you have any info on the samples complexity? If it is very complex with a high numer of proteins spanning a large range of abundancies you might need to thing about prefractionations which than clearly requires more starting material.
3. What type of equipment is available for your quantitative proteomics study. (ie. an old LTQ orbitrap platform or an Orbitrap Exploris 480 hybrid Orbitrap system).
In the end it is important to consider several factors including the desired statistical power, effect size, and variability in relation to the three points listed above.