Dear Arun,
The it works is to weigh a lot more to the given assignment not really just fix. So, even though you provide pre-assignment, if the data are not happy, theres a chance I-PINE is in trouble. Also, if your spectra do not align perfectly, the pre-assignment should be referenced to the N-HSQC. If theres some deviation, and another peak in the overlapped region is smaller euclidean distance to the given pre-assignment, that will also cause problems. If this is to deposit your chemical shifts to the bmrb, please just use export in the NMRFAM-SPARKY (two-letter-code es, or NMRFAM - Universal NMR formats - Generate NMRSTAR 3.1) not I-PINE for that. I-PINE generates probabilistic assignments with multiple choices, so thats not what you want upload to the archive. Since the question is related to the I-PINE, I would appreciate if you use I-PINE user group because some of people may already suffer from massive emails.
Best,
Woonghee