Adding an additional biomt distorts the coordinates. Therefore, in the sassie-web version, the process_biomt is comment out (sassie_web_utils and scaffold builder). Also need to rigorously debug this as it will add new segments etc. that must be documented and explained clearly.
Notes from DWW:
Off the top of my head:
The original coordinates should always be retained (identity matrix is skipped in transformation stage)
Answer (a) - there is an algorithm that chooses the 'next' segment name - should be the same as used when splitting chains in PDBs (i.e. when there is a protein + ligand in chain X). The idea here is that BIOMT are used to create an NMER from a monomer.
As I recall multiple BIOMT should produce multiple copy chains from the original coordinates using that transform. Which should be safe in the sense that no operations are chained. Thinking about allowing user added BIOMT in this context seems like madness to me now and I can't remember why I did it (or at least why I did it so early).
and my response:
Using biomt as “the way” to build extended structures isn’t really the way I want that functionality to be implemented. I think it will be easier to make a VMD/JSMol interface available and have users make duplicates of bits and move them by hand (or by inputing operations) while they “see” the results of their handiwork rather than entering in matrix elements in a GUI only.
Adding an additional biomt distorts the coordinates. Therefore, in the sassie-web version, the process_biomt is comment out (sassie_web_utils and scaffold builder). Also need to rigorously debug this as it will add new segments etc. that must be documented and explained clearly.
Notes from DWW:
Off the top of my head: The original coordinates should always be retained (identity matrix is skipped in transformation stage) Answer (a) - there is an algorithm that chooses the 'next' segment name - should be the same as used when splitting chains in PDBs (i.e. when there is a protein + ligand in chain X). The idea here is that BIOMT are used to create an NMER from a monomer. As I recall multiple BIOMT should produce multiple copy chains from the original coordinates using that transform. Which should be safe in the sense that no operations are chained. Thinking about allowing user added BIOMT in this context seems like madness to me now and I can't remember why I did it (or at least why I did it so early).
and my response:
Using biomt as “the way” to build extended structures isn’t really the way I want that functionality to be implemented. I think it will be easier to make a VMD/JSMol interface available and have users make duplicates of bits and move them by hand (or by inputing operations) while they “see” the results of their handiwork rather than entering in matrix elements in a GUI only.