Why the transformation matrix u and t are all zeros?

[zzz14]

I tried to align two proteins using easy-search and get the transformation matrix by --format-output 'u,t', but I got results like "0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000 0.000,0.000,0.000".

[milot-mirdita]

They shouldn't be all 0.
Can you please post the steps you executed so we can try to reproduce this.

[zzz14]

I put the attached two pdb files (github doesn't allow .pdb extension, so I changed the extension to .txt when uploading here) under the tmp folder. Then, I run the command

6fje_B_protein.txt
4x6k_A_protein.txt

/path/to/foldseek/bin/foldseek easy-search tmp/6fje_B_protein.pdb tmp aln tmpFolder --format-output 'query,target,alnlen,mismatch,qstart,qend,tstart,tend,cigar,bits,u,t'

This is the result I got

6fje_B_protein	6fje_B_protein	624	0	1	624	1	624	624M	6105	0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000	0.000,0.000,0.000
6fje_B_protein	4x6k_A_protein	660	271	25	535	1	475	7M4D4M1D5M1I15M5I5M7D9M1D5M7I5M1D11M14D10M5D6M4D6M2D6M21D3M42D5M27D26M10I6M5D9M10I3M3I9M1D13M4I7M1I4M2D2M16I8M10I15M7I6M4I6M1I2M2I13M6D4M2D4M1I7M4D5M16I9M6I4M9I9M34I12M19I3M7I11M1I14M2I3M3I3M6I7M	56	0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000	0.000,0.000,0.000

[hughhigin]

If it's helpful I just tried downloading and running the commands (my foldseek version is ca58f9b) and my output looks normal:

aln.txt

I tried installing the latest version (0dd4b7f) and it produces the same bug

aln_v2.txt

So it appears to be a bug introduced more recently, at least after june.

[martin-steinegger]

Thank you for reporting. @sooyoung-cha could you please have a look. It seems your filter is overwriting the u,t for complexes.

[martin-steinegger]

I have fixed this issue. Could you please try the newest version?

[zzz14]

It works now. Thank you very much!