Adding focal points to the map

Interesting paths and connections through the data.

Adding focal points to the map

Postby chaomei » Fri Aug 21, 2009 7:57 pm

Looks good. I will explore the site closely.
Perhaps a few points of interest could be shown on the map to attract people's attention to some specific spots, e.g.
where the Spanish flu, the Mexican flu, etc. And how does the map change weekly by showing either a sequence
of maps or an incrementally updated map (each week has increasingly darker color, for example, so one can seek
new additions of sequences with more ink).
Given a white background and light gray basemap, red or dark blue colors, or other colors with shading would give
shaper contrast.
Chaomei
chaomei
 
Posts: 3
Joined: Fri Aug 21, 2009 7:53 pm

Re: Adding focal points to the map

Postby donpellegrino » Fri Aug 21, 2009 9:02 pm

Thanks, I added a few points of interest to the map by coloring the circles for the respective proteins.

Green: H1N1 for 2009 (Mexican Swine Flu)
Code: Select all
WHERE  S.serotype = 'H1N1'
AND    T.strain LIKE '%2009%';

Red: H1N1 1918, H2N2 1957 and H3N2 1968.
Code: Select all
WHERE  S.serotype = 'H1N1'
AND    T.strain LIKE '%1918%';

WHERE  S.serotype = 'H2N2'
AND    T.strain LIKE '%1957%';

WHERE  S.serotype = 'H3N2'
AND    T.strain LIKE '%1968%';

Blue: Sequences that code for PB1-F2
Code: Select all
WHERE  G.gene = 'PB1-F2';

A problem with this simple coloring approach is combining sets. I'll try using alpha blending to combine the red with the blue where those strains encode for PB1-F2.

The temporal analysis will be more challenging. Recalculating a new layout each week will take a lot of computation and it may be disorienting since the landmarks and positions of proteins will shift with each run of the layout algorithm. I'll try using the given full layout and hiding nodes for multiple views depending on their time value. There is also an issue of selecting which time value to look at. NCBI provides a year attribute which appears to represent the year of infection for the virus. There are also update and create dates for the record itself at the granularity of a day.
Attachments
outliers.png
A possible mutation of the H1N1 2009 toward a red cluster.
outliers.png (51.09 KiB) Viewed 614 times
donpellegrino
 
Posts: 14
Joined: Wed Aug 19, 2009 1:52 pm

Re: Adding focal points to the map

Postby chaomei » Fri Aug 21, 2009 9:14 pm

With LGL, is it possible to freeze a set of nodes, while allowing the rest of nodes move freely? If that can be done,
you could freeze old sequences, and use the existing layout to accommodate newly arrived nodes.
chaomei
 
Posts: 3
Joined: Fri Aug 21, 2009 7:53 pm

Re: Adding focal points to the map

Postby chaomei » Fri Aug 21, 2009 9:16 pm

The new attached map is much more interesting. As least viewers would know, the green and red dots should be somehow more
interesting than others in the background.
chaomei
 
Posts: 3
Joined: Fri Aug 21, 2009 7:53 pm

Re: Adding focal points to the map

Postby donpellegrino » Fri Aug 21, 2009 9:29 pm

Unfortunately LGL does not appear to offer a feature to freeze the coordinates for a set of nodes. Since it is an iterative algorithm it should be possible to modify the code with a new starting point that includes nodes with fixed positions. I'd have to dig into the LGL code more deeply to see how this would be implemented. It would certainly be a useful feature.
donpellegrino
 
Posts: 14
Joined: Wed Aug 19, 2009 1:52 pm


Return to Analyses

Who is online

Users browsing this forum: No registered users and 1 guest

cron