Leuven2012:Stemmatology
From IntereditionWiki
Contents |
Phylogenetic computation tool/ Web Service
Group Members: Armin, Tara, Joris, Doug
Wednesday, 11.01.2012
morning
~ Assessment of existing services and online-tree-generators from bio-informatics
- http://molbiol-tools.ca/Phylogeny.htm, http://itol.embl.de/
- How do they implement it, what are their input/output formats
- => For a simple webservice from scratch in 2.5 days, input, output can be limited but GUI should already include all formats/computations
- Input format: aligned table structure (later on tei and nexus)
- output format: Newick (later on visual rendering)
- computations: Neighbor Joining [later on Parsimony, ... Münster]
~architecture
- server should use java for computation
~usable programms already written
- integrated distance matrix computation from tablestructure
- algorithms on the web (NJ)
~next steps: setting up server; setting up "hello world" webservice without functionality; plugging in and developing java-progs
afternoon
~computation
- creating a distance matrix from alignment table (Java)
- computing NJ tree in Newick format from that (Java)
- readiing SplitsTree
~architecture
- making a webservice by comnbining resources
- alignment table input -> either character or distance matrix -> compute tree from it e.g. in newick format -> compute graph and render graph
Thursday, 12.01.2012
- creating various algorithms and implementing them in Java:
create Distance Matrix from collation table, convert this into Nexus - format;
Friday, 13.01.2012
- Split out Tara's Text::Tradition::Stemma functionality into a standalone utility, webserviced, that takes a CollateX-style JSON alignment table and outputs a character matrix suitable for plugging into Phylip pars etc.
- Made a little web service here for this; a user should POST to it with form parameter 'alignment', and the JSON string as its value. The response is a JSON string with key 'matrix', whose value is the character matrix in a single string.
- implemented a conversion in Java from collation table to JSON String, agreement on format of table (#LACUNA for LACUNA); creating Java conversion from table to Newick tree(unfinished)