################################################### Archives of the omnipathdb.org web service contents ################################################### For reproducibility it is important to have all data and tools used in the analysis clearly defined. Each time we update the content of the web service we deposit here an archived content in order to maintain access for older versions of data. From the file names you see the time period when the data was served by the web service and also the query type (interactions, enz_sub, complexes, annotations, intercell). The files labeled `recent` contain the current content of the web service. ******* History ******* 24 May 2023 +++++++++++ - Fixed an issue that has been introduced in the 18 May update: dataset assignments for the "mirnatarget", "lncrna_mrna" and "small_molecule" datasets were missing hence these datasets returned no interactions - New custom edge attribute in CollecTRI: "sign_decision" 18 May 2023 +++++++++++ - Full details about interaction provenances in the new "evidences" column - New parameters "evidences" or "fields=evidences" - Interaction datasets are assigned based on the provenances of the interaction instead of simply based on the resource names - New custom edge attribute in CollecTRI: "tf_category" 5 April 2023 ++++++++++++ - Full contents update from the original resources - New gene regulatory network dataset: CollecTRI - New annotation resources: InterPro, Lambert2018, CytoSig 13 January 2022 +++++++++++++++ IMPORTANT: ---------- From now on, at each update all data are retrieved directly from the original resources. Before, only certain updates meant content update, while other updates meant only addition of new resources or features. Since the 20 Nov 2021 update, all updates come with fully updated and re-processed contents. A large number of resources in OmniPath update their data much less often (like every couple of years) or are not maintained at all, some since longer than a decade ago. The data from these resources is supposed to be identical across OmniPath releases. However, minor changes can be expected even in the data from these resources, due to improvement of the processing methods in OmniPath and due to changes in additional data used for the processing (such as ID translation data). - Since the previous update DoRothEA confidence levels were missing, resulting issues at data retrieval and filtering by confidence levels. Now the confidence levels are in place again, the normal behaviour has been restored. - miRecords and miRTarBase interactions in the miRNA-target network dataset were missing from the previous update, due to an issue in their input methods; these issues have been fixed. - New interaction type: `small_molecule_protein`, corresponding to the new network dataset `small_molecule`. Still small, featuring only three resources (SIGNOR, CancerDrugsDB and Cellinker). - New annotation resources: CancerDrugsDB, CellTypist, PanglaoDB and PROGENy. 20 November 2021 ++++++++++++++++ - Complete contents update and fresh build of all databases - New optional field in the `interactions` query: `extra_attrs`. This field contains resources specific, miscellaneous properties of the interactions in a JSON encoded format. Functions in the latest OmnipathR help to extract specific attributes from the JSON encoded column. The extra attributes can be accessed by using the `fields=extra_attrs` HTTP GET parameter. - The `dip_url` mandatory field from the `interactions` query has been removed as the `extra_attrs` include the DIP interaction identifiers under the key `DIP` -> `id`. - New ligand-receptor network, annotation and intercell annotation resources: scConnect, Cellinker, CellCall. 21 June 2021 ++++++++++++ - The files in the archive are compressed to save some disk space and bandwidth - The interactions expanded from the cofactor table of CellChatDB now labeled as `CellChatDB-cofactors` - The self-interactions (loops) were missing from the previous build of the interactions database - An error was fixed which resulted HTTP 500 for certain queries which should have returned empty tables instead 10 June 2021 ++++++++++++ - New protein complex resource: hu.MAP2 - New annotation resource: HumanCellMap 29 March 2021 +++++++++++++ - New option for interactions query: loops - include self interactions. - Fixed an issue with CellPhoneDB processing which resulted ~250 ligand-receptor interactions missing. This error had no effect on the annotations and intercell annotations. 30 January 2021 +++++++++++++++ - The CellChatDB ligand annotations were missing from the intercell database, we fixed this error 27 January 2021 +++++++++++++++ - The CellChatDB receptor annotations were missing from the intercell database, we fixed this error 21 January 2021 +++++++++++++++ - New network resources: Wojtowitz2020, CellTalkDB, CellChatDB, connectomedb2020, talklr - New annotation resources: CellTalkDB, CellChatDB, talklr - New intercell resources: CellTalkDB, CellChatDB, talklr - New complex resource: CellChatDB - Ramilowski2015 ligand-receptor interactions: excluding a few unconfirmed records which were previously included in OmniPath by mistake - Interactions data properly shows secondary resources (e.g. PhosphoSite_ProtMapper means the data is originally from PhosphoSite and came into OmniPath via ProtMapper) - Intercell data: resource name UniProt_keywords changed to UniProt_keyword - SIGNOR transcriptional regulation: only directed interactions - Improved filtering of gene regulatory interactions - The `resources` query filters the resources according to the `license` parameter - The license information is included in the `resources` query - Fixed behaviour of the `datasets` field in the `interactions` query - Handling duplicate field names - Content-length headers - `resources` field in `interactions` and `enzsub` queries - The server automatically restarts upon crash (max 15 min outage after each crash) - Serving the full annotations database is denied due to performance issues; instead the full database is available as a static file here 21 September 2020 +++++++++++++++++ - Fixed an issue in the last update: the interaction signs in the DoRothEA dataset were missing 9 September 2020 ++++++++++++++++ - We updated the DoRothEA dataset so it should be identical to the recent version in the DoRothEA git repo & R package 17 August 2020 ++++++++++++++ - Fixed: since the introduction of licenses the rendering failed when using query type synonyms - The `consensus_score` in the intercell query changed to be the number of resources (integer) instead of num. of resources / total num. of resources in the category - The `sources` parameter at the `interactions` query works again 11 August 2020 ++++++++++++++ - New column in the intercell query: `consensus_score`; it is the number of resources supporting a composite record normalized by the total number of resources contributing to each category; for non composite records it is NaN 3 August 2020 +++++++++++++ - Introducing the `license` parameter: possible values are academic (default), nonprofit, forprofit, commercial; this parameter ensures the resources contributing the returned data all fit the legal requirements of the user. Its main purpose is to help corporate users to meet the legal requirements. Academic users (which is the default setting) get more or less the total contents of our databases. - All databases have been rebuilt, however in most of the cases without updating the data from the original resources - New network and complex resource: KEGG-MEDICUS - New annotation resource: KEGG-PC (KEGG pathway annotations from PathwayCommons) 21 July 2020 ++++++++++++ - New resource in the intercell database: ICELLNET 19 May 2020 +++++++++++ - New network resources: Baccin2019, EMBRACE, ICELLNET, iPTMnet, iTALK - SignaLink3 network and annotation data updated - New annotation resources: Almen2009, Baccin2019, EMBRACE, CellCellInteractions, GPCRdb, ICELLNET, iTALK, MCAM, SignaLink_function, TCDB, UniProt (5 types of annotations: family, keyword, location, tissue, topology) - CSPA annotation contents updated - Complete redesign of the `intercell` database: the content is better curated, much more categories are available and the columns in the table represent clearly the attributes of the categories - Fixed many minor bugs in the server to ensure the correct processing of parameters (certain combinations of these resulted emtpy response, missing data or error) 17 Feb 2020 +++++++++++ - New enzyme-substrate and network resource: KEA - References in interactions and enzyme-substrate data are labeled with the source database, e.g. PhosphoSite:21465790 - New column in interaction and enzyme-substrate data (included on demand): `curation_effort`, showing the number of resource-reference pairs 16 Dec 2019 +++++++++++ - LRdb: new resource in annotations and intercell annotations - Some minor issues fixed in annotations: in annotations_summary the label was present in the list of possible values of some factor type variables; we made sure at all resources we have SwissProt IDs whereever it's possible - Interactions: at some interactions the list of PubMed IDs was redundant, we fixed this 10 Dec 2019 +++++++++++ - MSigDB has been added to annotations - An issue in the annotations_summary query has been fixed: in the previous version all combinations of source-label pairs were listed even if the label did not exists in the given resource - A new query type `resources` returns a JSON with comprehensive information about all the resources in all query types; at the moment it is limited to the datasets but it will be extended soon with information about the release dates, licenses, publications, URLs and many other things 27 Nov 2019 +++++++++++ - New enzyme-substrate interaction resource: ProtMapper - Annotations: subcellular location and secretome data from the Human Protein Atlas - Transcriptional regulation: not only DoRothEA but also original resources included: ABS, ENCODE, HTRI, ORegAnno, PAZAR, SIGNOR - A bug affecting UniProt IDs only a few interactions has been fixed (some UniProt IDs had an isoform suffix, e.g. P00533-1); as far as we know this affected very few UniProt IDs from PDZbase, phospho.ELM and IntAct. - Secondary resources (original resources imported via an other resource) are labelled with the resource we import in OmniPath, e.g. IntAct_CellPhoneDB means an IntAct interaction what we got via CellPhoneDB; at the moment this kind of resources are CellPhoneDB, DoRothEA, MIMP and ProtMapper. 18 Oct 2019 +++++++++++ - Complex components separator `_` instead of `-` - For records obtained from an intermediary resource the resources distinguished by a postfix e.g. for IntAct interactions coming via CellPhoneDB we state `IntAct_CP` instead of simply `IntAct` - Fixed wrong separators in resource name lists for the secondary resources mentioned above (sometimes there were commas and multiple resource names stuck together) - DEPOD enzyme-substrate interactions: input method fixed these were missing before but now present again in the data 04 Sep 2019 +++++++++++ - More details in annotations from DGIdb and kinase.com - Fixed processing of fields from COSMIC and CellPhoneDB 26 Aug 2019 +++++++++++ - New annotation resources (e.g. Human Protein Atlas, Cancer and Biological Pathway Associations Database [CPAD], Cancer Single-cell State Atlas [CancerSEA], etc) - Complex-protein interactions - `annotations_summary` and `intercell_summary` queries: a summary about the annotation fields and all possible values - Fixed: some TF-target interactions from DoRothEA were missing - Fixed: little formatting errors in some of the `annotations` - Fixed: minor inconsistencies in the `intercell` dataset 26 Jun 2019 +++++++++++ - 2 new extra datasets: pathway extra (pathwayextra), ligand- receptor extra (ligrecextra) - new literature curated pathway resource: Adhesome - majority consensus direction and sign (3 new columns in the interaction query) - new query type: `intercell` - pathway annotations and many more annotations from various databases - complexes represented in one line per each - an important bug fixed which previously filtered some interactions by accident (https://github.com/saezlab/pypath/issues/80); other minor bug fixes 25 Apr 2019 +++++++++++ - Updated content (fresh download) for most of the databases - New query types: annotations and complexes 16 Nov 2018 +++++++++++ - TF Regulons updated and PubMed references for this dataset have been added 14 Jun 2018 +++++++++++ - updated content: all resources re-downloaded from original sources - new server module provides more options in query interface - content upgraded to have 4 datasets: OmniPath, extra kinase-substrate interactions, miRNA interactions and transcriptional regulation from TF Regulons (DoRothEA) - homology translated interactions for mouse and rat provided ==vv== [ Archives below you find in directory `old` ] ==vv== Aug 2017 ++++++++ - additional enzyme-substrate interactions have been added Nov 2016 ++++++++ - content with OmniPath as it has been presented in the paper - in addition extra enzyme-substrate interactions (kinaseextra)