// structureToSequence uses structural fragments to categorize a segment
// of alpha-carbon atoms, and adds the corresponding residues to a
// corresponding sequence fragment.
func structureToSequence(
lib fragbag.StructureLibrary,
chain *pdb.Chain,
nullChan chan seq.Sequence,
seqChans []chan seq.Sequence,
) {
sequence := chain.AsSequence()
fragSize := lib.FragmentSize()
// If the chain is shorter than the fragment size, we can do nothing
// with it.
if sequence.Len() < fragSize {
util.Verbosef("Sequence '%s' is too short (length: %d)",
sequence.Name, sequence.Len())
return
}
// If we're accumulating a null model, add this sequence to it.
if nullChan != nil {
nullChan <- sequence
}
// This bit of trickery here is all about getting the call to
// SequenceCaAtoms outside of the loop. In particular, it's a very
// expensive call since it has to reconcile inconsistencies between
// SEQRES and ATOM records in PDB files.
limit := sequence.Len() - fragSize
atoms := chain.SequenceCaAtoms()
atomSlice := make([]structure.Coords, fragSize)
noGaps := func(atoms []*structure.Coords) []structure.Coords {
for i, atom := range atoms {
if atom == nil {
return nil
}
atomSlice[i] = *atom
}
return atomSlice
}
for start := 0; start <= limit; start++ {
end := start + fragSize
cas := noGaps(atoms[start:end])
if cas == nil {
// Nothing contiguous was found (a "disordered" residue perhaps).
// So skip this part of the chain.
continue
}
bestFrag := lib.BestStructureFragment(atomSlice)
sliced := sequence.Slice(start, end)
seqChans[bestFrag] <- sliced
}
}
// StructureBow is a helper function to compute a bag-of-words given a
// structure fragment library and a list of alpha-carbon atoms.
//
// If the lib given is a weighted library, then the Bow returned will also
// be weighted.
//
// Note that this function should only be used when providing your own
// implementation of the StructureBower interface. Otherwise, BOWs should
// be computed using the StructureBow method of the interface.
func StructureBow(lib fragbag.StructureLibrary, atoms []structure.Coords) Bow {
var best, uplimit int
b := NewBow(lib.Size())
libSize := lib.FragmentSize()
uplimit = len(atoms) - libSize
for i := 0; i <= uplimit; i++ {
best = lib.BestStructureFragment(atoms[i : i+libSize])
if best > -1 {
b.Freqs[best] += 1
}
}
if wlib, ok := lib.(fragbag.WeightedLibrary); ok {
b = b.Weighted(wlib)
}
return b
}