Atlantica gp25 (in PECAAN) is showing high-probability full-length hits to prokaryotic Pleckstrin homology (PH)-like domain proteins (e.g. 3HSA_D, Shewanella protein YP_926556.1, type example). Other HHPred hits with similar scores are to the same domain structure in eukaryotic proteins, where the class was originally identified (e.g. 2N23_A, TFIIH p62, canonical example of PH domain). There are dissimilarities in the lengths of some of the disordered regions between structural sub-domains, and identity is low, but conservation of the structure is much better than homology of the sequence. At 104 aa, Atlantica gp25 is similar in length to a known sub-group of bacterial PH domain proteins typified by 3HSA_D but is shorter than a previously identified group of phage-associated PH domain proteins (~180 aa). AlphaFold indicates the characteristic 7-strand beta "sandwich" core + C-terminal alpha helix of this protein class (pic attached).

I'm not sure whether PH-like domain meets the criteria to be added as a call. While PH-like domain is a structural class, it is a fairly specific 3' structure, and there is a function associated - canonically these proteins associate with membranes (lipid binding) and assemble other proteins at their location (protein-protein interactions). What do you think?

Xu et al. 2010. Bacterial Pleckstrin Homology Domains: A Prokaryotic Origin for the PH Domain,
Journal of Molecular Biology 396(1): 31-46. https://doi.org/10.1016/j.jmb.2009.11.006 (https://www.sciencedirect.com/science/article/pii/S0022283609013576)
1.19Mb