Link to this post | posted 23 Aug, 2018 18:10

cdshaffer

Annotation of phage Gilgamesh, gene 43 (27658-28101) in PECAAN. This prediction shows 99.4% prob, with ~90% query and subject coverage HHPRED hit to MutT/NUDIX type nucleotide pyrophosphohydrolase. There is an approved term for "MazG-like nucleotide pyrophosphohydrolase". According to this paper (The Journal of Biological Chemistry 286, 30691-30705) the nucleotide pyrophosphohydrolases fall into "structurally different superfamilies, such as the MutT-related hydrolases (Nudix) (α + β), dUTPase (all-β), dITPase (Maf/HAM1) (α/β), all-α-NTP pyrophosphatases (MazG), and phospho-ribosyl-ATP pyrophosphatase (HisE)." The MazG type are a superfamily made up of all alpha helices as can been seen in this PDB crystal structure of the eponymous MazG protein 3CRA. The MutT (also called Nudix type) show a kind of sandwich with a beta sheet between alpha helices like 4KYX I would propose adding either "MutT-like nucleotide pyrophosphohydrolase" to keep a parallel structure to the other entry or just have a generic "nucleotide pyrophosphohydrolase" at the bottom of the list for those enzymes that are not in the MazG superfamily. Edited 23 Aug, 2018 18:11

Link to this post | posted 20 Jul, 2018 19:17

cdshaffer

To answer your question on Fe+2 binding, I could not find any in vivo studies but this result is a crystal of an "iron soaked" version of the protein complete with Iron bound as expected and a discussion of the conserved Asp channels that could allow ions to flow into the cavity. I agree on every single point you made. I really don't think there is a good annotation here. I think maybe the best option here is "do no harm" so we will proceed with Yaboi and annotate this gene NKF. We can always change it later.

Link to this post | posted 20 Jul, 2018 18:43
cdshaffer	DNA processing protein A is on the list, it is officially: DprA-like DNA processing chain A two rows below the red Ku in the second column

Link to this post | posted 13 Jul, 2018 16:13
cdshaffer	For me when the top of the "add a gene window" is cut off it is because I have the zoom level set to magnify the page. This happens a lot as I typically have my pages default to 150% zoom or so. The tiny 3-line menu in the top right corner can be used to see and or change the current zoom level.

Link to this post | posted 13 Jul, 2018 16:00

cdshaffer

Getting quite a few HHPRED hits to a large collection of related crystal structures for Yaboi_draft 37 (15,584 to 15183 Rev) results are available on PECAAN as gene 36. Gene is member of pham 3874. The top 50 or so HHPpred hits are all 99.9% probability and >99% coverage with titles that include the term DPS or DPS-like. There are then a large number of 99.8% probability >99% coverage hits to ferratin from various species from Humans to M tuberculosis. Looking into the literature a bit DPS is short for "DNA protection during starvation protein", these proteins are a family of proteins that have a ferritin-fold, show DNA binding activity, and whose expression is induced during starvation or other sources of oxidative stress, see this paper for more info on the original e coli protein and this on a M smegmatis DPS. The issue is what term to use for these protein. "DPS protein" seems too simplistic and uninformative. On the other hand, it would make it really easy for annotators to match hhpred hits to the proper annotation term. About 1 in 10 hhpred descriptions do include the full "DNA protection during starvation protein" so we could use this longer term. A few descriptions have "ferritin DPS family protein" which I like didactically as it links ferritin an important protein conserved across all kingdoms. So maybe following the recent trend in approved terms, would the best term be "ferritin-like DPS protein"?

Link to this post | posted 03 Jul, 2018 19:54

cdshaffer

Phage Justbecuase (cluster BM, available on pecaan) gene 252 starting at 149782 has a more than 50 HHPRED hits all with >99% probability and 85%-95% coverage with a variety of descriptions. Examples include "RNase III inhibitor", "Appr-1-p processing domain", "O-acetyl-ADP-ribose deacetylase", and most commonly something about "Macrodomain protein". It turns out there is a large body of literature on "macrodomain" proteins which have been shown to be involved in a wide variety of biological functions across all kingdoms (this wikipedia article is pretty good at summarizing). Interestingly, one of the well characterized prokaryotic uses of this domain has been shown to be involved a toxin/antitoxin system that use UDP-riboylation of DNA. This domain was found and annotated previously in phage Trina gene 253. Here full annotation in the genbank file is "ADP ribose-1-P processing protein, antitoxin" while phagesdb appears to drop the last part and just has "ADP ribose-1-P processing protein". Looking over quite a few of the published structures and enzymatic functions of many of the HHPREd hits, the common thread appears to be enzymes that break/create bonds between an ADP-ribose moiety and something else, be that a protein (see the PARPS), a nucleic acid (as in the DART/G toxin anti-toxin), a phosphate group (in the APPR1-p processing proteins), nicotinamide (see the NAD(+) ADP-ribosyltransferases) , or an acetyl group (see the O-acetyl-ADP-ribose deacetylases). I really don't like the generic term "macro domain" even though most of the literature uses that term. Instead I would propose the term "ADP-ribosyltransferase" which I think does a good job of summarizing all the different activities that have been associated with various macrodomain proteins. While the toxin/antitoxin angle is intriguing I am not sure there is sufficient evidence to support including this biological role as part of the annotation without supporting evidence from something other than HHPRED or BLAST alignments.

Link to this post | posted 29 Jun, 2018 17:22

cdshaffer

Online aragorn call the reverse strand tRNA from 88542-88627 as a Pyl amino acid and an anti-codon of cta (which pairs with the amber stop codon UAG). Pyl (pyrrolysine) is one of the noncanonical amino acids which have been described but are fairly rare. A pubmed search of Pyrrolysine in streptomyces only found examples of using a system cloned out of an archaea into streptomyces for use in amber supression. online tRNA scan calls this with an infernal score of 55.7, type undet and anti-codon of NNN. So, are we calling these specialized amino acids and anti-codons if called by Aragorn 1.2.38? Or should we go to a less specific designation of a tRNA with unknown aa and anti-codon? Edited 29 Jun, 2018 20:46

Link to this post | posted 29 Jun, 2018 15:12

cdshaffer

We have a gene in the BM streptomyces phage Satis [gp211 with annotated start @129464], with around 40 HHPRED hits with > 99% prob and > 98% coverage with short descriptions of "cold shock protein" and one hit of similar quality to an "RNA chaperone". Investigation into these cold shock proteins suggest they also act as chaperones. See this JBC paper. I think either a generic "chaperone" or a more specific "RNA chaperone" would summarize the results here. For the recent submission we took the more conservative route of just "chaperone".

Link to this post | posted 22 Jun, 2018 16:45

cdshaffer

Yes, you are correct, in every case I have worked with lately Pecaan is giving the codon in the downloaded notes (i.e the reverse complement of the anti-codon reported by tRNA-scan). I always change the final notes back to the anti-codon before creating the genbank submission file. Since Pecaan reports the codons in uppercase, I make the anti-codons lower case which helps me keep track if that step has been completed and matches the examples from the online documentation. In genbank the "tRNA-Thr(ggt)" info ends up in the "Notes" field. There is no official standard for that field in the official documentation here. So I think the important thing is to have everything from the SEA-Phages use the policy as stated in the online docs linked above.

Link to this post | posted 18 Jun, 2018 15:17
cdshaffer	I thought the same thing but no luck. The intein is entirely within the second (in order of transcription) gene and the upstream gene appears to no longer be expressed.