Hi, Everyone, In protein databases such as NCBI and others. The same protein often has many different entries (different accesion numbers deposited by different authors at different time). To avoid the redundancy, I need to pick one entry for each protein to work with. What is the criateria to pick the "best" one (the most accurate)? Any rule of thumb for a programmer? Thanks a lot. Ren (rxwang1@aol.com)