I tried it since last 5 months and now going to complete it soon, but the problem is that i dont know how to get most important protein and how to discuss it in the discussion. I also dont know how and on what bases to pick the most important item of cellular component, molecular function, biological process and kegg pathways.
We don't know if OP has tried to talk/discuss the situation with the professor. @arjahejo has stated above that he comes from an underdeveloped country and is probably not used to the idea of learning as an interactive process. Unfortunately in most underdeveloped countries knowing more than your mentor is a considered a professional hazard so one gets used to doing things as asked, even if they don't make sense. Critical thinking/problem solving abilities (essential for a PhD) that students in developed countries get exposed to early in their undergraduate years are sadly absent most elsewhere.
@arjahejo: Tell us if you have given regular updates to your professor along the way. If you have not done this then it may be the reason why you find yourself to be in the situation you are in now. If you have been talking with the professor and are not receiving any guidance then you should seriously consider switching labs. PhD is a long term relationship and if you can't work with the mentor you will face bigger troubles down the road.
i am very near to it, please guide me further.
You haven't told us what you have done and the reasons behind those choices. What kind of an experiment was this that yielded the data you have. There may be no right answers for many scientific problems. Your success will be decided by the path you took, the results you got and a scientific explanation that you were able to formulate to fit those results/observations.
I must say that your professor sucks and you must report him (some universities allow this). Your mentor should be a "mentor", explain to you what is the purpose of the analysis, what is the biological context and if he/she is not capable to teach you the necessary skills, to support you with classes/courses/workshops.
Relate to your question, a Gene Ontology (Enrichment) analysis simply reports if a group of genes belongs to some particular categories based on statistics like "what is the probability of having 10 genes belonging to the same GO class based on a test group". Tools like AmiGO or DAVID, Panther, GOrilla, etc. can do the test, simply use your list of genes as input and select categories by a P-value threshold keeping classes that are statistically overrepresented in your sets. But the question about what is important in biological terms it is only answered knowing the experiment, what are the conditions? which metabolic pathways are expected to be modified? why 40hr? etc.