| | 1 | Some notes that I have about this: |
| | 2 | |
| | 3 | PSICQUIC is planned as a common web service API for all |
| | 4 | participating databases. Matthias Oesterheld demonstrated an initial |
| | 5 | schema for this tool, based on the HOBIT experience of document |
| | 6 | style web services. As previously discussed, this will be a leveled |
| | 7 | service with Level 1 supporting simple queries such as get all |
| | 8 | interactions by organism (TaxID), by interaction property, by |
| | 9 | experiment property or by protein list. All attending databases |
| | 10 | would be capable of supporting a cpath implementation of this model, |
| | 11 | others may need a wrapper web service. It was decided to change |
| | 12 | protein list to interactor properties and extend the search to |
| | 13 | encompass short name, aliases and cross-reference. Searching by |
| | 14 | TaxID would encompass all children of the term used. Negative |
| | 15 | interactions would not be searched in initial implementations. Level |
| | 16 | 2 could incorporate more complex queries and feature properties |
| | 17 | searched by both !InterPro cross-reference or MI type. |
| | 18 | |
| | 19 | Returns would be in a document, the maximum length of section |
| | 20 | returned could be set by the operator, by generating a start index |
| | 21 | and length. An additional method on the web server could then |
| | 22 | retrieve the desired number of interactions in batches. A test |
| | 23 | procedure could pick out specified entry numbers allowing the user |
| | 24 | to skip large datasets. The possibility of adding a flag to the |
| | 25 | query to give the option of a URL at which a compressed complete |
| | 26 | file could be accessed was also discussed. |
| | 27 | |
| | 28 | Complex queries would not be too difficult to establish, however |
| | 29 | may prove too great a server load. It was decided that both levels |
| | 30 | would be published simultaneously, with Level 2 being completely |
| | 31 | backwards compatible. Implementers may then decide which level to |
| | 32 | install. The first query will send a user query. The return will |
| | 33 | include a query ID and the number of results. A retrieval function |
| | 34 | will detail query ID, count, start index and length. Query send may |
| | 35 | be adjusted to size of client server. Standard error messages will |
| | 36 | be incorporated. An initial implantation should be available late 2005. |
