Error makes no sense while use FunFolDes

[jeteveux]

I encountered a puzzle bug with FunFolDes.It keeps saysResidue 10 not found in pose!I checked the input and confirmed there are no missing resi.I set the argument"error_on_out_of_bounds_index" to false to try to locate the bug,but it didnt work out.
I have attached the relevant files ,cloud anyone kindly provide some advice.
FFD.zip

[roccomoretti]

So the error is coming about because FunFolDes is chopping up the protein into pieces, and when it operates on the isolated pieces it looks for the 10th residue and can't find it.

The reason it can't find the 10th residue is because the "donor" Index selector is only selecting regions of 6aa, which means those pieces are only 6aa long. The reason why it's looking for the 10th residue is because you told it to looks at lengths of 3, 6, 10, 16, & 20 in the NubInitioMover's Nub Segment's "editable" setting. Removing 10, 16 & 20 from that setting will fix that issue.

However, you will then run into a "The number of template pieces ( 4 ) has to be twice the number of Nub regions ( 2 )". (That later number has already been multiplied by two, so the issue is that you only have one Nub region.) There are four template pieces because you have two discontinuous segments in your template_motif_selector (a template segment is generated on each side of the continuous stretched of residues in the template_motif_selector). The "easy" way of fixing that would be to simply double the number of <Segment/> subtags in the NubInitioMover's Nub subtag (and be sure to specify the order correctly).

Now whether these changes actually do what you want them to do is an open question. -- I'd double check the FunFolDes documentation and the examples to double check things are set up the way you expect them to be.

PS. Note that there's a fair amount of "[ ERROR ] couldn't find fragment to insert!!" errors when you run your protocol as-is. This is due mostly to your chain A being 825 residues, whereas the protein you're trying to refold around your loops is only 97 residues. Because of the details in the way the refolding works, this large asymmetry in "fixed" and "foldable" residues results in difficulties finding a fragment which passes the tests. Note that you can try adding the -abinitio:apply_end_bias false command line option, which will change how the refolding works to avoid it. (Though that will also change how the refolding works.) Trimming the binding partner can also be of help.