In Supplementary Figure 9, it is demonstrated that fTM-align significantly speeds up the alignment of larger structures. However, using the -fast flag for multimer alignment has no effect on the calculation time or the alignment results. Below is the code to replicate this, I'm using the current version of US-align from this repository.
wget https://files.rcsb.org/download/7NYH.cif
wget https://files.rcsb.org/download/7NYU.cif
time ./USalign 7NYU.cif 7NYH.cif -mm 1 -ter 0 -fast | head -16 | tail -3
time ./USalign 7NYU.cif 7NYH.cif -mm 1 -ter 0 | head -16 | tail -3
gives
# fTM-align
Aligned length= 2195, RMSD= 0.83, Seq_ID=n_identical/n_aligned= 1.000
TM-score= 0.47471 (normalized by length of Structure_1: L=4615, d0=18.82)
TM-score= 0.99666 (normalized by length of Structure_2: L=2195, d0=14.28)
real 0m12.052s
# TM-align
Aligned length= 2195, RMSD= 0.83, Seq_ID=n_identical/n_aligned= 1.000
TM-score= 0.47471 (normalized by length of Structure_1: L=4615, d0=18.82)
TM-score= 0.99666 (normalized by length of Structure_2: L=2195, d0=14.28)
real 0m12.052s
Is there a reason why fTM-align doesn't work for multimers?
Hi,
In Supplementary Figure 9, it is demonstrated that fTM-align significantly speeds up the alignment of larger structures. However, using the
-fast
flag for multimer alignment has no effect on the calculation time or the alignment results. Below is the code to replicate this, I'm using the current version of US-align from this repository.gives
Is there a reason why fTM-align doesn't work for multimers?
Thanks!