Re: Implementing DALI in 3D Coffee

[Peter Hyungjun Yoon]

Hi,

I apologize if this is a novice question; 

I saw on structural aligners portion of your documentation that DALI is listed. Is there any more details on how I could combine DALI and 3D align for refining of my MSA? I have both programs installed locally. I am wondering how TMalign could also be implemented. For the latter, is it simply the case that I must have TMalign installed localled, and its path added to my bashrc profile?

Thank you for making such a powerful program. I’ve been getting frustrated at dealing with divergent sequences with clear structural homology, especially in the post-alphafold era where we have so many accurate structures available. 

Best,

Peter

[Cedric Notredame]

Hi Peter,

I am not sure about dali. We have an interface with dalilite. I have not tried it in a long time (when the server was still at the EBI), but if everything has been kept the same, then I guess you could try with the latest dalilite.pl : http://ekhidna2.biocenter.helsinki.fi/dali/README.v5.html

For the other: all you need is having TMalign compiled and installed on your system. In practice, we have often found that a combination of TM-align and sap (i.e. -method=sap_pair, TMalign_pair) gives the best results. Also bear in mind that if your sequences have disordered regions, you may also want to combine all this with a sequence based procedure (-method=sap_pair, TMalign_pair, proba_pair). In theory you also could throw in dali_pair considering the above caveat.

Have fun!

Cedric