Closed ValWood closed 7 years ago
If we use JBrowse for this, there is a dependancy here: https://github.com/GMOD/jbrowse/issues/772
This is what I have so far as a replacement for the genome browser image. It's a bit ugly because I'm not artistic.
This is what would be shown on the pom1 gene page.
Targets inferred from the MF section are shown in black E.g.Rga4, Cdr22 (these represent direct interactions) Targets inferred from the BP section are shown in blue E.g.Tea1, Dis2 (these represent direct OR indirect interactions, but they shouldn't be really upstream or indirect) Targets inferred from the phenotype section are shown in red E.g. Imp2, Mac1 (these could be direct, same pathway, indirect or upstream)
I wanted to incorporate the Biogrid Physical Interactions too but then (I guess) we'd have to lose the "directionality" of the graph.
The "other" category on the RHS is for anything that cannot be shoehorned into the GO slim categories associated with pom1.
Actually, I think it would be great to incorporate the BioGrid physical intearctions as well. That's the only way of getting things that associate with the same structures as your gene of interest, even if they don't act on each other directly enzymatically.
You could stick these in the middle so that at the top you have things targeting your gene of interest. At the bottom you have targets of your gene. And in the middle you then have things that are either targets, targeting, or just associated with the same complex, as your gene of interest.
For clarity I think genes should get omitted from the middle if they are already shown as a target, or targeting.
Sorry the diagram still looks like British beach huts on speed.
(If there is now webservice, its only a viewer + a pdb file)
Of course it could be more simply a static structure image like: http://www.rcsb.org/pdb/explore/explore.do?structureId=5IKK in the basic info section..... but the rotating 3d view would be so cool!
I might prefer this for structure (simpler) http://biojs.io/d/bio-pv rotatable manually
Discussion at the planning meeting makes me think that a section summary at the top of the page is a good way to GO (I will make a better mock up for this)
At the plannign meeting I suggested thte the GO summary could look, for example, like this:
cyclin-dependent protein serine/threonine kinase activity involved in regulation of mitotic cell cycle phase transition, cytokinesis, conjugation with cellular fusion, mitotic sister chromatid segregation, microtubule cytoskeleton organisation, DNA replication, DNA recombination, and DNA repair
Molecular function x annotations with x direct target genes linked to x processes x binding partners Biological process x annotations Cellular component x locations x complexes
(the first part is essentially a "slim view" and the second part summarizes the main types of data)
Clicking a link would take you to the page section with the current page section
The over arching topic is supposed to be how to summarize page sections at the top of the page.
I plan to break this down once discussed further, its on my list.
I still think we should do some cool stuff at the top of the page. For later...
I've added a separate issue about JBrowse: #443
Section 1 "Basic information"
What we have now:
Gene standard name systematic ID synonyms Product Genomic location (complete transcript and CDS) characterisation status (published, inferred, etc) Feature type (protein,ncRNA etc) Name description(s) Products size (aa and Mwt) contig map showing context links to adjacent genes