What is a good method to reconstruct a protein's 3D structure from a pair-wise distance matrix of residues?

I think the question is self-explanatory. I know of this method but it seems that it uses contact maps, which actually use a certain cut-off. I don't have a cutoff based contact matrix but a matrix with actual distances.

  • $\begingroup$ You have distance between alpha carbons or beta carbons (or other atoms of the residues)? Usually a threshold is set because below certain measure it can't be and above certain threshold it doesn't mean a contact between the residues. Do you have any other information besides the distance matrix, like the sequence of the protein? $\endgroup$ – llrs Nov 26 '18 at 8:55
  • $\begingroup$ @llrs They are only carbon alpha distances and I want to reconstruct the backbone with carbon alphas and no more atoms.No more information, just the distances between carbon alphas for each residue. $\endgroup$ – Wise Nov 26 '18 at 16:17
  • $\begingroup$ You could use the multidimensional scaling approach, but I suspect it might propose invalid angles between the backbone. It would be possible to use it as a starting point before checking the angles between the backbone. But I don't know if there is any software to do this (it is a strange problem to have), or either you know the sequence or the sequence and the structure, but just the distance is strange. $\endgroup$ – llrs Nov 27 '18 at 9:50

Finding coordinates of N points that preserve (as well as possible) pair-wise distances from an NxN matrix is called multidimensional scaling.

But I don't enough about the context of your problem to tell if/how it is applicable.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.