What is TMPDB?

The importance of prediction methods of transmembrane segments in proteins is now increasing in the scene of genome sequencing projects, because they can be used to give clues to assignment of structure and function to genes. A problem in developing the prediction methods is the lack of 3D structure data of membrane proteins. There are only a few entries in PDB. But in the case of membrane proteins, topology (2D structure) can be determined experimentally, such as chemical modification of residues, cleavage by enzyme, gene fusion method (alkarine phosphatase etc.), even if the 3D structure of them are not available.
TMPDB is a collection of membrane proteins whose topology have some experimental evidence. They are collected from published papers. It may be used to construct test/training data set of transmembrane segments or topology prediction methods.

Format Description of TMPDB 

ID   Identifier; same as in Swiss-Prot if exists in it.
OR   Organism.
ME   Membrane in which the protein exists.
NO   Entry number in TMPDB.
AC   Accesion number in other databases, e.g.  Swiss-Prot.
DE   Definition; same as in Swiss-Prot if exists in it.
DT   Date of adding the entry.
RA   Authors of reference.
RL   Reference.
NS   Number of transmembrane segements.
SS   Structure of transmembrane semgents, i.e. alpha or/and beta.
TP   Topology.
FT   Feature; Location of transmemrane segments in the sequence.
SQ   Sequence  and  its length.
//   Terminator line of a entry.

Reference:

Shimizu, T. & Nakai, K., "Construction of a membrane protein database and an evaluation of several prediction methods of transmembrane segments.", Genome Informatics Workshop 1994, pp.148-149, Universal Academic Press, Tokyo, Japan, 1994. - Abstract [ps| pdf]
Last Modified 20 April 1998