Hi,
this is a feedback (or rather very vague idea) from one of our collab partners from the clinical side.
When looking at assembly graphs, they try to "navigate" the genome in terms of chromosomes, chromosome bands, big structures like telomeres etc.
In essence, their mental model is still very linear in the sense of an IGV browser view where you can click on a chromosomal region and jump there. How could a similar thing be achieved in strangepg?
My view: surely, having some sort of reference information is not a must for strangepg, but also a fairly common use case. To provide some analogy: in RTS games, you commonly have a mini-map on screen for approximate navigation. Could one think of a similar concept by creating such a coarse-grained overview panel somewhere in the corner of the screen that contains a sorted list of identifiers / tags taken from the graph file? For example, assume that all nodes carry a tag identifying the chromosome band. The overview panel lists those tags (specified by the user "create global view from tag" or the like) in a sorted order and upon clicking on such a tag, strangepg jumps to the location in the graph (if necessary: to the location with highest density for such tags; one could cycle through all clusters of there are several...) that is identified by this particular tag.
Hope that is somewhat clear as an idea for future developments.
Best,
Peter
Hi,
this is a feedback (or rather very vague idea) from one of our collab partners from the clinical side.
When looking at assembly graphs, they try to "navigate" the genome in terms of chromosomes, chromosome bands, big structures like telomeres etc.
In essence, their mental model is still very linear in the sense of an IGV browser view where you can click on a chromosomal region and jump there. How could a similar thing be achieved in strangepg?
My view: surely, having some sort of reference information is not a must for strangepg, but also a fairly common use case. To provide some analogy: in RTS games, you commonly have a mini-map on screen for approximate navigation. Could one think of a similar concept by creating such a coarse-grained overview panel somewhere in the corner of the screen that contains a sorted list of identifiers / tags taken from the graph file? For example, assume that all nodes carry a tag identifying the chromosome band. The overview panel lists those tags (specified by the user "create global view from tag" or the like) in a sorted order and upon clicking on such a tag, strangepg jumps to the location in the graph (if necessary: to the location with highest density for such tags; one could cycle through all clusters of there are several...) that is identified by this particular tag.
Hope that is somewhat clear as an idea for future developments.
Best,
Peter