My research area is Data mining and my work on that is regarding "Protein Secondary Structure Prediction using support vector machine". I’m using CB513 &RS126 dataset.I run the sequence without using PSI-BLAST. But I got wrong result because of insufficient features and then I have read many papers regarding the sufficient and insufficient feature. I understand that, we have to run the sequence in PSI- BLAST for much iteration and see for which sequence we get high accuracy and use that particular sequence for structure prediction, is it right? I thought to confirm what I have understand. The output from the BLAST is completely different from the input I gave, Next doubt is we are giving only amino acid sequence (primary sequence), how we can get secondary structure for result in sequence. How to view the PSSM from PSI-BLAST. What is scoremat? why do we need to BLAST the amino acid sequence? In Basic Local Alignment Search Tool,how to upload PSSM? My research is right or wrong. Please tell me the correct way. How can I precede my research work?
Please kindly reply I have less time to complete my research work