See proposal from 2010 workshop.
Also Binary data support in Dalliance.
Briefly: there’s a pretty good alignment data model, but the query syntax currently has a few issues which limit its use for genomic alignments. Most significantly, there’s not a good way to fetch a subset of the alignment based on sequence coordinates (e.g. “give me all alignment blocks overlapping the segment 22:30000000,30100000 from human NCBI36”).
Can we extend DAS to give better support for more interactive client navigation? Things like next/previous feature buttons.
Customized html labels/pop ups? (JW)
Improvements, Schema Driven? (JW, RJ))
For search functionality (or in general?) (GS, AJ, RJ, JW, LG)
Should the DAS registry say list all sources and just give information on their validity rather than archiving sources. Should clients do all they can to accommodate poor data sources or should they link to validation data and encourage correct data source compliance?(JW)
Also: how aggressively should the registry check for and prune dead datasources? Should there be some expectation of “maximum reasonable response time” for a DAS request? (TAD)
As simple as it can seem, binning strategies for the maxbins attribute are not evident. Although the binning startegy is very dependant on data type, it would be nice to have a small set of standard simple binning algorithms so maxbins can be easily added to new DAS sources. (BG)