Link to this post | posted 26 Jun, 2020 17:09
welkin	Hi Kirk, Can you stitch the lysin back together in silico and get really good BLASTN matches to other lysins? That would be the most compelling evidence for an intron, other than the HNH (the HNH in the gap is already pretty compelling).

Link to this post | posted 18 Jun, 2020 17:00
welkin	We've published a paper on the genomics of some of the Cluster K phages, including describing a start-associated sequence that can help you decide which start is the real start for genes that have this sequence. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0026750 Edited 01 Dec, 2023 04:49

Link to this post | posted 02 Jun, 2020 21:33
welkin	Thanks Greg!

Link to this post | posted 22 May, 2020 15:43

welkin

so with 101 I'd be inclined to root around a little in the ORFs in the forwards directions to see if there is an HNH in there somewhere. HNHs are frequently found associated with tRNA clusters, and sometimes give false positives with glimmer/genemark with a prediction in the opposite strand. If you don't find an HNH anywhere, I have no problem with the delete reverse 101 decision. 10bps should be fine for an overlap between 109 and a downstream tRNA.

Link to this post | posted 09 May, 2020 17:07
welkin	Hi Matt, I moved your post because the "cluster tips" are reserved for reports of cluster-specific oddities rather than a place to ask questions. In this case, if you have the data for the more specific assignment, by all means use it! The other designation is for instances when you can't always tell. BEst, Welkin

Link to this post | posted 06 May, 2020 21:36

welkin

hi Ellen, no. There should be only one capsid maturation protease per phage, sometimes a MuF-domain will be fused onto this protease. There can be multiple proteases– just not multiple ones of this specific kind. Does that help? Make sure that anything you are going to call as a protease has an actual protease domain match when you use HHpred. Sometimes things get mislabeled and then those labels get propagated through the database. We are working on the cleanup. Best, Welkin

Link to this post | posted 06 May, 2020 18:54

welkin

Hi Fred, Thanks for your detailed documents; it makes helping you much easier. First: You have two genes. The reasons you should not delete 42 is it has better coding potential than 43 and it is present in at least one other phage (Fowlmouth– as you pointed out). So: now that you are keeping 42; you have to use a start that does not capture all the CP of 43, as that would lead to a substantial overlap with 42 without a good reason (good reasons include high quality function alignments). The best start is the most conserved start— again, the one annotated in the other phage and selected by Glimmer; visible in Starterator. Both of those things absolutely outweigh the RBS score (both scores are pretty bad, honestly, so it is hard to consider it at all). I hope that helps. Best, Welkin Edited 06 May, 2020 23:05

Link to this post | posted 09 Mar, 2020 12:33

welkin

Thanks Chris. This should be "endolysin". According to my most recent conversation with Graham about this, the Lysin A and B designations arose because there are two in the Mycobacteriophages. Other phages, which just have one, have historically just used the term "endolysin". So we will be changing all of the non-Mycobacteriophage lysins to "endolysin" except in cases in which we can identify a clear lysin B (I think this only applies to some of the Gordonia phages). I haven't gotten to updating the Streptomyces yet.

Link to this post | posted 03 Mar, 2020 23:40
welkin	I think with those sequences I would go with no slip for now.

Link to this post | posted 02 Mar, 2020 14:36
welkin	Hi Shallee– sounds like a good decision to me.