WormBase December 2015: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
Line 184: Line 184:
===Semi-automated curation using the Textpresso information retrieval system===
===Semi-automated curation using the Textpresso information retrieval system===


We also routinely employ the Textpresso information retrieval system for semi-automated curation of GO Cellular Component and Molecular Function annotations.
We also employ the Textpresso information retrieval system for semi-automated curation of GO Cellular Component and Molecular Function annotations.


===Computational annotation strategies===  
===Computational annotation strategies===  

Revision as of 09:13, 9 December 2015

Overview:

Staff

Paul Sternberg, PI, WormBase, GO [8%; 0% funded by GOC]

Juancarlos Chan, Developer, WormBase [25%; 25% funded by GOC]

James Done, Developer, Textpresso [40%; 40% funded by GOC]

Kevin Howe, Project Manager, WormBase - EBI [  ; 0% funded by GOC]

Ranjana Kishore, Curator [25%; 10% funded by GOC]

Yuling Li, Developer, Textpresso [30%; 20% funded by GOC]

Jane Lomax, Curator, WormBase ParaSite [  ; 0% funded by GOC]

Hans Michael Mueller, PI, Textpresso [75%; 50% funded by GOC]

Daniela Raciti, Curator [10%; 0% funded by GOC]

Kimberly Van Auken, Curator [100%; 75% funded by GOC]

Annotation Progress

WormBase GO Annotation Statistics as of December 1, 2015

Manual annotation statistics are summarized in Tables 1 - 3.

Total number of unique manual annotations: 39290 (+43.3% from 2014)

Total number of genes with manual annotations: 6762 (+44.2% from 2014)

Table 1: Summary of C. elegans Manual Biological Process Annotations

Numbers refer to total number of annotations; annotations in parentheses represent annotations with extensions.

Annotation Group IMP IGI IDA ISS TAS IEP IPI IC NAS ISM ND IBA
WormBase 7649 (367) 3104 (78) 1105 (23) 324 (1) 111 285 (58) 51 51 (10) 32 2 3 0
UniProt 801 (229) 155 (94) 126 (9) 190 25 (2) 14 2 (2) 5 104 0 65 0
CACAO 18 1 3 0 0 0 0 0 0 0 0 0
GOC 59 14 261 122 17 0 4 2 10 0 0 379
BHF-UCL 11 0 0 0 0 4 0 0 0 0 0 0
MGI 5 0 0 0 0 0 0 0 0 0 0 0
HGNC 0 0 0 4 0 0 0 0 0 0 0 0
GO_Central 2 0 0 4 0 0 0 0 0 0 0 8119
ParkinsonsUK-UCL 10 (4) 5 (2) 9 2 (1) 0 0 0 0 0 0 0 0
Totals 8555 (600) 3279 (174) 1504 (32) 646 (2) 153 (2) 303 (58) 57 (2) 58 (10) 146 2 68 8498


Table 2: Summary of C. elegans Molecular Function Annotations

Numbers refer to total number of annotations; annotations in parentheses represent annotations with extensions.

Annotation Group IMP IGI IDA ISS TAS IPI IC NAS ISM ND IBA ISO
WormBase 152 (6) 33 1658 (188) 650 (1) 47 1298 (2) 15 5 4 63 0 2
IntAct 0 0 0 0 0 1989 (52) 0 0 0 0 0 0
UniProt 37 (1) 2 119 (2) 181 21 272 (3) 4 51 0 126 0 0
CACAO 1 0 7 0 0 0 0 0 0 0 0 0
GO_Central 0 0 0 0 0 0 0 0 0 0 5531 0
HGNC 0 0 0 2 0 0 0 0 0 0 0 0
ParkinsonsUK-UCL 0 0 4 0 0 2 (2) 0 0 0 0 0 0
Totals 190 (7) 35 1784 (190) 833 (1) 68 3561 (59) 19 56 4 189 5531 2


Table 3: Summary of C. elegans Cellular Component Annotations

Numbers refer to total number of annotations; annotations in parentheses represent annotations with extensions.

Annotation Group IMP IGI IDA ISS TAS IPI IC NAS ISM ND IBA
WormBase 9 (3) 0 5818 (738) 345 26 140 46 6 4 8 0
GO_Central 0 0 0 0 0 0 0 0 0 0 4927
UniProt 24 (9) 1 246 (31) 196 16 0 19 50 0 118 0
GOC 36 11 22 6 0 0 3 0 0 0 47
MGI 0 0 14 0 0 0 0 0 0 0 0
HGNC 0 0 0 8 0 0 0 0 0 0 0
BHF-UCL 0 0 7 0 0 0 0 0 0 0 0
Reactome 0 0 0 0 5 0 0 0 0 0 0
CACAO 0 0 3 0 0 0 0 0 0 0 0
Totals 69 (12) 12 6110 (769) 555 47 140 68 56 4 126 4974


Table 4: Summary of C. elegans Computational Annotations

Based on WormBase Release WS250

Total number of genes with Phenotype2GO-based Annotation: 6,809

Total number of genes with InterPro2GO-based Annotation: 11,282

Type of Annotation IEA
Phenotype2GO Mappings - WormBase 42,666
IEA/InterPro2GO - WormBase 35,082

Methods and strategies for annotation

Curation methods

Literature curation

Curation of the primary literature continues to be the major focus of our manual annotation efforts.

Over the past year, WormBase has begun a topic-based approach to curation in which curators focus curation efforts on one or more biological topics, or processes, for each release cycle. Topics over the past year have included the endoplasmic reticulum and mitochondrial unfolded protein responses, innate immunity and defense response, and Wnt signaling pathways (see below).

Semi-automated curation using the Textpresso information retrieval system

We also employ the Textpresso information retrieval system for semi-automated curation of GO Cellular Component and Molecular Function annotations.

Computational annotation strategies

Our computational annotation strategies include mapping genes to GO terms using InterPro domains and mapping genes to Biological Process terms based upon mappings between terms in the Worm Phenotype Ontology (WPO). Beginning with the WS246 WormBase release, these Phenotype2GO-based annotations will include phenotypes based upon genetic variations as well as RNAi experiments. Results from automated methods are generated anew with each WormBase database build to reflect any changes in the underlying reference genome sequence and/or gene models.

Curation strategies

Priorities for annotation

Selection of genes for annotation is guided by several criteria:

  • Annotation of gene sets involved in specific biological processes as part of WormBase's coordinated topic-based curation process
    • Topics annotated to date: Unfolded Protein Response (ER and mitochondrial), innate immune response, defense response to pathogen, and Wnt signaling
  • Genes identified in Textpresso-based curation pipelines
  • Re-annotation of genes affected by changes to the ontology, e.g. cilia biology, ubiquitination, enzyme regulator activities
  • Publication of newly characterized genes
  • C. elegans genes orthologous to human disease genes

Presentations and Publications

Papers with substantial GO content

  • Van Auken K, Schaeffer ML, McQuilton P, Laulederkind SJF, Li D, Wang SJ, Hayman GT, Tweedie S, Arighi CN, Done J, Müller HM, Sternberg PW, Mao Y, Wei CH, Lu Z. BC4GO: A Full-Text Corpus for the BioCreative IV GO Task.Database (Oxford). 2014 Jul 28;2014. pii: bau074. doi: 10.1093/database/bau074. Print 2014. PMID: 25070993, PMCID: PMC4112614
  • Mao Y, Van Auken K, Li D, Arighi CN, McQuilton P, Hayman GT, Tweedie S, Schaeffer ML, Laulederkind SJF Wang SJ, Gobeill J, Ruch P, Luu AT, Kim JJ, Chiang JH, Chen YD, Yang CJ, Liu H, Zhu D, Li Y, Yu H, Emadzadeh E, Gonzalez G, Chen JM, Dai HJ, Lu Z. Overview of the Gene Ontology Task at BioCreative IV. Database (Oxford). 2014 Aug 25;2014. pii: bau086. doi: 10.1093/database/bau086. Print 2014. PMID:25157073, PMCID: PMC4142793
  • Huntley RP, Harris MA, Alam-Faruque Y, Blake JA, Carbon S, Dietze H, Dimmer EC, Foulger RE, Hill DP, Khodiyar VK, Lock A, Lomax J, Lovering RC, Mutowo-Meullenet P, Sawford T, Van Auken K, Wood V, Mungall CJ. A method for increasing expressivity of Gene Ontology annotations using a compositional approach. BMC Bioinformatics. 2014 May 21;15:155. doi: 10.1186/1471-2105-15-155. PubMed PMID: 24885854; PMCID: PMC4039540.

Presentations including Talks and Tutorials and Teaching

Poster presentations

Other Highlights

Ontology Development Contributions

  • Terms Added to the Ontology in 2014:
    • lysosome-related organelle
    • gut granule
    • gut granule lumen
    • gut granule membrane
    • peptidyl-proline 4-dioxygenase binding
    • tail spike morphogenesis
    • regulation, positive, negative anterograde synaptic vesicle transport
    • positive, negative regulation of pharyngeal pumping

Annotation Outreach and User Advocacy Efforts

  • Kimberly Van Auken continues to serve on the GO-help rota.
  • Kimberly Van Auken assisted with migration of content to the new GO website.

Annotation Advocacy

  • Kimberly Van Auken is participating in bi-weekly calls on development of the LEGO curation model and accompanying curation tool, Noctua.

Other Highlights

WormBase Data Models and Software

  • Progress Reports - Juancarlos Chan and Kimberly Van Auken have written a new script for reporting our manual annotations statistics. This script reports the number of annotations per contributing group according to evidence code and also reports the number of annotations with annotation extensions.
  • WormBase GO Annotation Model - Kimberly Van Auken, Kevin Howe, Paul Davis, and Daniela Raciti collaborated on development and testing of a new GO annotation model for WormBase. The model will allow for full incorporation of annotation extension data into WormBase, as well as additional annotation details and new IEA annotations from the UniProt-GOA group. The new model and accompanying data will be included in WormBase Release WS237.
  • Phenotype2GO Annotation Pipeline - Kimberly Van Auken, Juancarlos Chan, and Kevin Howe revised the Phenotype2GO-based annotation pipeline to incorporate both RNAi- and genetic variation-based phenotypes to increase coverage for these types of annotations. Concurrently, they also changed the curation pipeline to allow for improved quality control by housing all annotations within a curation tool hosted at Caltech.
  • WormBase Ontology Browser - Raymond Lee and Juancarlos Chan, in collaboration with the GO software team at Berkeley, developed a new ontology browser for WormBase, WOBr using the new AmiGO2.0 software. The browser allows for searching across multiple ontologies used in WormBase, including GO, WPO, and the worm Anatomy and Life Stage Ontologies.

Text Mining and Textpresso Central

  • Kimberly Van Auken and Yuling Li developed a new support vector machine (SVM) document classifier for a subclass of the molecular function ontology: catalytic activity. This SVM is now included in the WormBase data flagging pipeline that also includes classifiers for macromolecular interactions, expression patterns, and RNAi- and variation-based phenotypes.
  • Monica McAndrews (MGI), Kimberly Van Auken, and Yuling Li are collaborating on a document classification pipeline to help MGI identify papers suitable for curation. Using training and testing papers supplied by MGI, we have developed an SVM classifier to distinguish mouse from non-mouse papers. The next step in the process will be to help MGI develop a pipeline for identifying mouse markers (genes) associated with experimental data in these papers.
  • Hans-Michael Muller and Yuling Li started developing a literature curation platform named Textpresso Central that enables curators to perform full text literature searches, view and curate research papers, train and apply machine learning and text mining algorithm for semantic analysis and curation purposes. The user is supported in this task by giving him capabilities to select, edit and store lists of papers, sentences, term and categories in order to perform training and mining. The system is designed with the intent to empower the user to perform as many operations on a literature corpus or a particular paper as possible. It uses state-of-the-art software packages and frameworks such as the Unstructured Information Management Architecture (http://uima.apache.org), Lucene (http://lucene.apache.org), and Wt (http://www.webtoolkit.eu/wt). The corpus of papers can be build from fulltextarticles that are available in PDF format (http://en.wikipedia.org/wiki/Portable\_Document\_Format) or NXML (http://dtd.nlm.nih.gov/). An extension for articles published in HTML (http://en.wikipedia.org/wiki/HTML) is planned.

Back to http://wiki.geneontology.org/index.php/Progress_Reports



Back to http://wiki.geneontology.org/index.php/Progress_Reports