if the crystal structure of any protein is unavailable,

then one can use the tools of homology modelling to

determine the structure. the logic is that a similar

structure arises becuse of a similar sequence of amino

acids. for homology modelling to be accurate an identity

match of 70% is desirable.basically, one does a FASTA

search of A.A sequences in the PDB database (A.A sequences

whose structures are known), does a CLUSTAL alignment to

check for conserved residues. then the structure of the

unknown A.A sequence is built up on the basis of the

structure of the best matches in FASTA and CLUSTAL by

programs like LLOOP and HHPRED. this structure can be

visualized in programs like DEEPVIEW,Protein Explorer

etc.