Release Notes - iRefWeb - Wodak Lab

Version 4.1 — 2011 December 15

• Added information about legacy IDs for interactions. See FAQ regarding legacy identifiers.

Version 4.0 — 2011 December 9

• Now using iRefIndex release 9.
• MI (MINT Inspired) scores have been calculated for interactions. LPR and HPR filters have been replaced with the MI (MINT Inspired) Score filter.
• Interaction detection method and type filters now propagate their counts to their parent terms. Thus selecting a term will also return all of its children.

Version 3.9 — 2011 September 08

• Search filter counts and the actual search results can now be by PubMed, in addition to interaction and protein.

Version 3.8 — 2011 August 17

• Simplified and reorganized statistics page. It now provides only a high level overview of the landscape of protein interactions. For more detailed queries, the filters on the search page are capable of providing data distributions for more rich and detailed custom queries.

Version 3.7 — 2011 August 11

• PubMed Detail tab has been renamed PubMed Annotations. It's functionality has been enhanced to include a summary table. Also, the detail table has sortable columns now. Finally, as has been requested by users, there is no longer any limit on the display. So regardless of the number of interactions and proteins observed in a paper, it's annotation details will be displayed.
• PubMed Reports has been removed; there did not seem to be much interest in this functionality which is also largely duplicated in the Search and PubMed Annotations pages.

Version 3.6 — 2011 July 28

• Search terms now emphasize the use of protein/gene aliases over full text for better matching. Further, the user can control to which protein the alias is mapped in iRefWeb.

Version 3.5 — 2011 July 12

• New search filters (Entrez Gene Pairs, and Protein Alias Pairs) added that allow you to enter either two sets of gene IDs or two sets protein aliases and see if any interactions exist between the two sets.

Version 3.4 — 2011 March 02

• The protein (interactor/show) pages have been updated to provide more information:
  • A list of the protein's binary interactors
  • A list of the protein's possible complexes (if any)
  • A list of the protein's self interactions (if any)
• Links to a protein's page can now be found the on the search results, and interaction pages.
• A protein's aliases have reduced to only those aliases that provide link-outs.
• Because the interaction page now links the participating proteins to their own page, the only details shown when a protein is expanded is how it was mapped for that interaction. More details are also given on the mapping process.
• If you have a (Entrez) Gene ID you can try accessing it directly using this URL
  http://wodaklab.org/iRefWeb/interactor/showForGene/5420
  replacing '5420' with your (Entrez) Gene ID of interest. Note this does not imply that all genes can be found in iRefWeb.

Version 3.3 — 2011 February 14

• Updated to use iRefIndex version 8.

Version 3.2 — 2010 December 21

• In addition to the two original download formats, a third has been added: a simplified version of the MITAB with just five columns and mostly unprefixed aliases:
  • iRef Interaction ID
  • primary alias A
  • primary alias B
  • A iRef ID
  • B iRef ID

Version 3.1 — 2010 December 16

• Search now can return results and filter counts by protein in addition to interaction.
• Searching via the protein alias filter has been improved to provide more focused results (annotations have been removed from the aliases in this field).

Version 3.0 — 2010 October 28

• Cleaner more compact layout of search page features and functionality.
• Dynamic search-filter counts with as well as options for how selected filters are applied (any, all, none).
• You can enter multiple PubMed IDs and protein aliases — these fields are now in the filter section along with uploading your (Entrez) gene ID list(s).
• Improved display of interaction evidence on the interaction/show page.
• New MITAB file used for downloading results:

  MITAB—this format is based on recent changes agreed upon by the PSI-MI working group in Turku, Finland 2009. For details of the file format please look here.

  PLEASE NOTE: These files are not have not been canocialized like your search results have been, so where in the event isoforms were mapped together in iRefWeb, they are considered as separate records in the MITAB file.

  These files contains more information in additional columns (54 columns in total) . Columns 1-36 are part of the new PSI-MITAB 2.6 format. Additional columns are added by iRefIndex to track provenance of the original data and record how the iRefIndex procedure consolidated data.

  Each line in this file represents a single source database record that supports either:

  • an interaction between two proteins (binary interaction) or
  • the membership of a protein in some complex (complex membership) or
  • an interaction that involves only one protein type (multimer or self-interaction).

  Multiple lines may present information that support some kind of interaction between the same set of proteins. These related lines may be grouped together using the key provided in column 35.

• Broken stats links on 'interaction/show' page fixed.

Version 2.3 — 2010 September 08

• Bug fix for download of search results in MITAB format (results incorrectly filtered/matched).

Version 2.2 — 2010 August 26

• Some filters changed from discrete steps (1,2,3, etc.) to ranges (1 or more, 2 or more, etc.). Filters affected: Number of Supporting PubMeds, Number of Database Records, Number of Annotating Databases.
• Small warning added to the Detection Methods and Interaction Type filters (click on the help icon beside the title).

Version 2.1 — 2010 July 23

• Clarifcation notes added to stats page/tab.
• New search filters:
  • whether or not the interaction was predicted or experimental
  • the count of distinct source databases annotating the interaction
  • the total number of source database records supporting the interaction

Version 2.0 — 2010 July 13

• Using iRefIndex version 7.
• Improved search results: now in a table format with sortable columns.
• New search filters, now with slider controls for custom filter settings.
• Interactome downloads are no longer limited to 50,000 records or less.
• Search results can also be exported as a list of interaction ids for use in OrthoNets.
• You can now upload a list of Entrez Gene IDs to search on instead of searching for just one ID at a time.
• This release includes data from BINDTranslation. This data is being treated as if it was a non-public database because it is from an early version. Future versions of iRefIndex/iRefWeb will include BINDTranslation as public data.

Version 1.0 — 2009 March

• iRefWeb 1.0 contains the first release of the iRefIndex dataset (iRefIndex 6) in which an additional isoform-consolidation step was introduced into the iRefIndex procedure. Starting from this release, every interacting protein is mapped to the canonical splice isoform of the corresponding gene whenever possible. This enables further consolidation and more reliable comparison of the interaction data across the source databases (see irefindex.uio.no/wiki/Canonicalization).
• iRefWeb 1.0 and iRefIndex 6 were used in the preparation of two upcoming publications:
  • Turner et al. "iRefWeb: Interactive analysis of consolidated protein interaction data and their supporting evidence", Database Journal, 2010 (in press) -- a detailed description of the iRefWeb resource, its many interactive features and several usage scenarios.
  • Turinsky et al. "Literature curation of protein interactions: measuring agreement across major public databases" (manuscript under review) -- a comparative study of current literature-curation practices that pertain to protein-interaction data.
• The original source-database releases used in iRefWeb 1.0 / iRefIndex 6 are listed at irefindex.uio.no/wiki/Sources_iRefIndex_6.0.