Membrane proteins are crucial for survival﹒They constitute the key components for cell–cell signal transduction﹐transport of ions or solutes across the membrane, and are crucial for recognition. Many methods predict membrane helices, but few predict membrane strands﹒The good news is that most methods for helical membrane proteins are available and have higher accuracy﹒Current prediction methods predict membrane helices for about 50%–70% accuracy, and 10%false prediction for the globular protein﹒The bad news is that developers have seriously overestimated the accuracy of their methods﹒ This research utilized neural network to predict membrane protein. The protein data were collected from PDB database. The standard data set contains 593 membrane protein sequences and 500 non-membrane protein sequences. Nine popular prediction tools on the internet were used to assess its accuracy. Then, we take the way of neural network to integrate these tools and to predict membrane protein more precise.
