Link to this post | posted 14 Jul, 2025 21:51

cdshaffer

There is a region in many AS phage (typically around 18 - 23 kb) where genemark and glimmer call different genes on different strands. These genes often have very large or complete overlap. This is unusual in the extreme and finding the proper strand is a known weakness of these ab initio gene finders. As such be very suspicious of genes which overlap and are on two different strands. To call both you need very strong evidence, ideally mass spec results showing translation of both frames. However extremely high quality HHPRED hits for both potential predicted protein products might also suffice. Without extremely strong support, review the evidence and pick one gene and delete the other following the rules as outlined in the annotation guides. Edited 14 Jul, 2025 21:53

Link to this post | posted 12 Jul, 2025 21:11
cdshaffer	database 605 was released with 3 new phage. The whole starterator reports for these phage have been added to the collection found here: https://wustl.box.com/v/Actino-phage new phage in 603: Bezza LookingGlass OrediggerDelux

Link to this post | posted 02 Jul, 2025 04:06
cdshaffer	database 603 was released with 4 new phage. The whole starterator reports for these phage have been added to the collection found here: https://wustl.box.com/v/Actino-phage new phage in 603: Wolfwood Pisa4 Demure Colossa

Link to this post | posted 19 Jun, 2025 20:29
cdshaffer	As of 2020 there is a new rule, you should have a lysin B if you are annotating a lysin A. If you cannot find a lysin B just annotate endolysin. See this discussion: https://seaphages.org/forums/topic/4656/

Link to this post | posted 18 Jun, 2025 21:18

cdshaffer

There is a protein in this cluster which has been annotated with two different functions: deoxycytidylate deaminase and nucleoside deoxyribosyltransferase. Check the pham for this protein: PhluffyCoco_CDS_54. These two enzymes appear to be related and have an ancient common ancestor but these two enzyme activities are quite different. Be wary of BLAST alignments for this function and focus on the top HHPRED alignments or 3D predictions. For several phages in the BS cluster the HHPRED evidence is clear these are nucleoside deoxyribosyltransferase.

Link to this post | posted 17 Jun, 2025 15:41
cdshaffer	for more discussion on this matter see this discussion: https://seaphages.org/forums/topic/5775/

Link to this post | posted 14 Jun, 2025 01:53
cdshaffer	Database 602 was released with 15 new phage. The whole starterator reports for these phage have been added to collection found here: https://wustl.box.com/v/Actino-phage New Phage released in 602: Chaewon DaRealMyers Ding FremontTroll Labradorite Mikronejon Montavir Peridot RegnumDei Sangak SassyC SuperSonics Trice Virizion Zarafa Edited 02 Jul, 2025 04:07

Link to this post | posted 03 Jun, 2025 17:18

cdshaffer

We had a recent putative RecA that had some curious results so am following up on the "How to call a RecA". Our protein had what appeared to be two significant differences to the standard model. One of the hydrolytic residues did not appear to be present and the C-terminal Mg binding domain was totally missing. See the attached for more details but the final result was that our protein was still a RecA once we investigated, and shows the limits of HHPRED and the idea that anything conserved is necessary for activity. 953Kb

Link to this post | posted 02 Jun, 2025 15:48

cdshaffer

Phage genomes are circular when inside the bacteria, and there are different mechanisms of how the phage goes from a linear genome in the phage particle to the circular form. Depending on the mechanism the genome assembly program may not be able to recognize the physical end of the genome and put it at base 1 of the assembly. If this happens base 1 will be somewhere in the middle of the contig. Thus for all phage you have to inspect the assembly and "rotate" it if necessary to get to the publishable form of the sequence. This is done on a case-by-case basis, and is almost always needed, however, sometimes it is not. So for a paper you need to know specifically for your phage, so best to email Dan directly and ask.

Link to this post | posted 13 May, 2025 15:02

cdshaffer

yes. The genome is going to circularize after entry, at which point there is only one copy of the repeat within the genome. This means that annotations of the linear genome will always have the possibility of this kind of quirk. In addition, if you annotated a partial gene at the start of the genome it would cause all kinds of issues for the all the computational checks and make handling the genome much more labor intensive, so the best approach is as you suggest. Annotate the copy that is the full intact gene at the end and do not annotate the partial gene at the beginning.