Hi Rosetta’s List,
Has anybody happen to know how the minirosetta threading protocols works?
Considering a structural alignment composed of several homologous and parent sequences, If I understood something, minirosetta is going to make “nstructure” models for each pairs of sequence&pdb of the alignment. So, the total number of models is gonna be “nstrucure” times the number of sequences&pdbs pairs. Is not clear, however, if minirosetta mix up and weight the structural information from each pdb or it makes “nstructure” models based on each pdb separately. Another question, is there any rule of thumb to select the number of parent structures/sequence for alignment? Is it make sense to use an alignment with tens of sequences?
Could someone, please, shed some light on these questions?