Biodiversity Informatics Projects Richard White Thoughts • Role of biodiversity data in bioinformatics – assisting with organising and retrieving bioinformatic (molecular) data – a separate area with different users (taxonomy, ecology, conservation, resource management …) • Demand from users for taxonomic and species diversity information on the Web • Pressure on the taxonomic community to deliver • Demand for more sophisticated use of available data: interoperability = online analysis, not just browsing Assembling biodiversity information sources Delivering species diversity information by • • • • assembling, merging & linking databases and publishing on the Web, with special emphasis on linking Issues in assembling and linking biodiversity information sources • Assembling a web-site (ERMS) • Assembling databases by merging (ILDIS) • Linking on-line databases through a gateway (Species 2000 and SPICE) • Onward links to related information • Checking the reliability of links (LITCHI) • Intelligent linking • Persistent identifiers Assembling species databases First of all, before we start merging and linking databases, let’s assemble a database from scratch: • ERMS (European Register of Marine Species) • Now at www.marbef.org/data/erms.php ERMS Incoming data • Approximately 100 separate lists for different taxonomic groups • Mostly compiled as spreadsheets • Scientific names, synonyms, geography (at least Atlantic or Mediterranean) • Some optional fields • Objective to create a book and a web-site, partially supported by a database List conversion was carried out in several stages: • Excel spreadsheets were exported to text files • Tab-delimited text files were imported into a clientserver database (MySQL) • Database queries results are passed through templates to generate either RTF (for the printed publication) or HTML (for the Web site) Variations on a theme • Fields may be combined or separated e.g. genus species authority date • Higher taxa may be: – repeated in fields of the species record – given once in separate preceding records in various different formats • Synonyms may be: – in a separate field of the species record, or mixed with other remarks, with various delimiters and separators – in separate records, linked by code or by name or even abbreviated – implied, e.g. Genus1 specname (Smith as Genus2) • Geographical information is often free text ERMS book page Osteichthyes: brief checklist Reptilia: full details Taxonomic hierarchy for Reptilia Merging versus linking • Merging databases to create a single larger database • Linking databases to create a distributed information system Merging species databases 1 The original databases are physically copied into Plants of a new combined Europe database. Plants of Africa 1 2 The user interacts with the new combined database. Plants of the World 2 Linking 1 The user interacts with an access system which does not itself contain data. 2 When the user requests data, it is fetched from the appropriate database. Plants of Europe Plants of Africa 2 Plants of the World 1 Assembling databases by merging Now we have some databases, let’s build a bigger one by merging: • ILDIS (International Legume Database and Information Service) ILDIS International Legume Database and Information Service • International collaborative project – 10 Regional Centres – 30 Taxonomic Coordinators • Its goals include – building, maintaining and enhancing the ILDIS World Database of Legumes – designing and providing services from it to users, including: • ILDIS LegumeWeb • via Species 2000 ILDIS World Database of Legumes v. 7.00 • Taxa • Species • Subspecies • Varieties 15,500 1,600 2,400 19,500 • Names • Accepted names • Synonyms 39,500 19,500 19,000 ILDIS’s data model: core data • A core taxonomic checklist, assembled from regional data sets and nearing completion, provides a consensus taxonomy - a unified taxonomic treatment or backbone on which other data can be hung • Various kinds of additional data may be attached to this backbone (see later) Features of ILDIS LegumeWeb We’ll look at examples of the use of LegumeWeb, to show a couple of features: • Two-stage access with “synonymic indexing” • A gateway to external information - “onward links” (direct species name links) to further sources of information User access to LegumeWeb: Step 1 • The user types in a name, which may be incomplete (or wrong!) • LegumeWeb responds by showing a list of the species names which fit the user’s specification User access to LegumeWeb: Step 2 • The user chooses one of the species names provided (which may be synonym or an accepted name) – In this example, the user chooses Abrus cyaneus (a synonym for Abrus precatorius) • LegumeWeb responds by showing a standard set of information about the chosen species Synonymic indexing • Automated synonymic indexing • synonym entered accepted name found (name taxon) • taxon found synonyms listed • Types of synonyms – Unambiguous – Ambiguous • pro parte • homonyms • misapplied names • In these cases an explanation is offered to the user Assembling databases by linking Now we have some biggish databases, let’s build something even bigger by linking databases together: • Species 2000 – SPICE – Species 2000 Europa Linking 1 The user interacts with an access system which does not itself contain data. 2 When the user requests data, it is fetched from the appropriate database. Plants of Europe Plants of Africa 2 Plants of the World 1 The Catalogue of Life (Species 2000) • An international collaborative project to provide access to an authoritative and up-to-date checklist of all the world’s species • A distributed array of Global Species Databases (GSDs) can be accessed through a Web gateway or Central Access System (CAS) • The array of GSDs provide an index to a further range of information about each species, using onward links (see later) • www.sp2000.org Species 2000 organisation Taxonomic hierarchy (or hierarchies) Species Global species databases (GSDs) and interim checklists: the species index Species information sources (SISs): regional faunas and floras, specialist or sectoral databases, web pages etc. interim GSD checklists SIS Architecture of Species 2000 User interface Data collector (CAS) Wrapper Wrapper Wrapper GSD GSD GSD Species 2000’s Common Access System • Species 2000 gives users a single point of access to GSDs • Access involves a two-stage search process similar to that used in LegumeWeb • In the second stage, the user sees a screen of “standard data” about a species The “standard data” This comprises the information about a species which Species 2000 wishes to provide: – Accepted name (with references) – Synonyms (with references) – Common Names (with references) – Family or other higher taxon – Geography – Comment – Scrutiny information – URL or URLs linking to further data sources for this species Need for communication • Different people are building the various components of the system: – GSDs – wrappers – CAS – user interface • We need to ensure they all have a common understanding of the data to avoid mistakes Common Data Model • We use a Common Data Model (CDM) – A definition of the information being passed to and fro – Human-readable, not machine-readable – Helps to manage complexity – Used to create specific machine-readable implementations for Corba (IDL), CGI/XML (DTD, XML Schema), Web Services, etc. What does the CDM look like? • It defines the input (“request”) and output (“response”) for six fundamental operations which the system needs to be able to carry out Request Types 0-6 – Type 0: Get CDM version supported by a GSD’s wrapper – Type 3: Get information about a GSD – Type 1: Search for a name in a GSD – Type 2: Fetch “standard data” about a chosen species – Type 4: Move up the taxonomic hierarchy (towards the root of the tree) – Type 5: Move down the taxonomic hierarchy (towards the species level) Spice CAS in use Screen-shots of an old version of the Spice system in use: Spice 1 CAS Onward links to related external data Species databases such as ILDIS and federated systems such as Species 2000 envisage providing links from their data to external sources of related data, socalled “onward links” • Example from ILDIS ... “Onward links” • The user may follow a hyperlink to some other data source for further information, not managed by ILDIS – In this example, the user chooses to go to W3Tropicos at Missouri Botanical Garden to see more information • In this way LegumeWeb acts as a gateway to other information about legume species LegumeWeb page with onward links Destination of an onward link Further information obtained Checking the reliability of links • Whether in – merging data sets to construct a species database like ILDIS, or in – linking from one data set to another, • it is necessary to ensure that the species concepts in the different databases do not conflict Example 1 Database A • Caragana arborescens Lam. [accepted name] Caragana sibirica Medikus [synonym] Database B • Caragana sibirica Medikus [accepted name] Caragana arborescens Lam. [synonym] Example 2 Database A • Caesalpinia crista L. [accepted name] Database B • Caesalpinia crista L. [accepted name] Caesalpinia bonduc (L.) Roxb. [accepted name] Caesalpinia crista L., p.p. [synonym] LITCHI project • We modelled the knowledge integrity rules in a taxonomic treatment • The knowledge tested is implicit in the assemblage of scientific names and synonyms used to represent each taxon • Practical uses include – helping a taxonomist to detect and resolve taxonomic conflicts when merging or linking two databases – helping a non-taxonomist user follow links from one database to another, in which the species may be differently classified Conflict display Outcome of LITCHI project • A prototype tool for merging checklists & checking integrity of individual checklists was implemented • In the Species 2000 Europa project, we are now creating a completely new second version with a view to allowing: – dynamic linking (so-called “taxonomically intelligent links”) – Presentation of “attached data” to be organised, merged and used to support conflict resolution “Intelligent” linking • The Catalogue of Life (Species 2000) is – not just a catalogue (which lists things) – it is an index (which points to things) • GSDs, and gateways to them such as the Catalogue of Life, can serve not only as catalogues of species but also as indexes giving access, potentially, to all species information on the Internet “Intelligent” linking • Species 2000 plans to provide links to take a user – from a species entry (from a GSD) – to further sources of information about that particular species (Species Information Sources or SISs) Species 2000 organisation Taxonomic hierarchy (or hierarchies) Species Global species databases (GSDs) and interim checklists: the species index Species information sources (SISs): regional faunas and floras, specialist or sectoral databases, web pages etc. interim GSD checklists SIS “Intelligent” species links • Given that it is possible to detect many cases of potential taxonomic conflict when linking species databases, how can such links be managed? • There are a number of choices in the ways links may be made and handled Cross-mapping So how can we make intelligent links work, especially in the difficult cases where a species in one database does not have an exact match in the other ? – One way is to create and maintain “crossmaps” which describe how one or more taxa in one resource (such as the Species 2000 index) relate to one or more taxa in another resource A dream A system for managing intelligent species links would maximise the potential of the plethora of species-based catalogues, indexes and rich species resources currently being assembled all over the world • Perhaps on the Web, as with the current Spice/Species 2000 prototypes • Or ... The Grid The Grid is often thought of as a new toy for particle physicists, with – very high bandwidth – distributed computational resources But it also provides opportunities for more structured and reliable access to data and information sources, using improved protocols with metadata – For example, access to such knowledge sources as these cross-maps Using biodiversity information resources • Helping Biodiversity Researchers to do their Work • Collaborative e-Science and Virtual Organisations Biodiversity analysis and modelling Scientists working with biodiversity information employ a wide variety of resources: • data sources • statistical analysis and modelling tools • presentation or visualisation software which may be available on various local and remote computer platforms. Examples of biodiversity resources Data sources: • Names: Species 2000 & ITIS Catalogue of Life • Data: GBIF, sequence databases • Geography: Gazetteers • Collections and distributions: BioCASE, MaNIS Analysis tools: • Statistical and multivariate analysis • Modelling • Visualisation Use of resources together Scientists frequently need to use several of these resources in sequence to carry out their research. Much effort is currently expended in • initially acquiring resources • installing and sometimes adapting them to run on the user’s own machine • converting and transporting data sets between stages of the analysis process Biodiversity research Biologists are working to understand the adaptation of organisms to their environmental niche, eventually by combining knowledge of all the levels of biological organisation • genome • transcription • proteome • metabolic pathways • cell • tissue • organ • individual whole organism • population • species • evolutionary pathways and to predict their interactions with their environment Workflows Resources are called into use in an appropriate sequence from an interactive workflow. The facility for scientists to be able to create their own workflows, without the need for regular assistance from computer scientists, is an essential part of the BDWorld system. Accessible tools for resource discovery and for workflow design, enactment and reuse are therefore required. For example Changes in distribution in response to climate changes brought about by global warming CSM: Climate-space modelling Modelling and predicting changes in distribution in response to climate changes such as those brought about by global warming An unreasonably brief explanation: • Get current distribution of a species (e.g. specimen records) • Get current or recent climate data for those localities • Calculate a model for the climate space the species can occupy • Predict the distribution the species would have in any specified climate (may be different to the climate used Example work-flow (Climate-space Modelling) Submit scientific name; retrieve accepted name & synonyms for species Retrieve distribution maps for species of interest Possibly different climate surfaces (e.g. predicted climate) SPICE Climate Climate surfaces Model of climatic conditions Localities Climate Climate where species is currently Space Modelfound Prediction Prediction of suitable regions for species of interest Base Maps Projection of predicted distribution on to base map Projection World or regional maps Triana screen-shots 1 Creation (design, editing) Triana screen-shots Triana screen-shots Triana screen-shots Triana screen-shots 2 Execution (enactment, run-time) Triana screen-shots Triana screen-shots Triana screen-shots Triana screen-shots And finally … Triana screen-shots Elements of the BDWorld system • What did the system have to do to make that example happen? Role of the work-flow engine • Create and edit a workflow – locate an appropriate resource – check interoperability – arrange any necessary transformations – record provenance of generated data sets • Execute a workflow, passing data sets to and fro • Create a log or ‘lab book’ for user Difficulties with resources • Finding the resources • Knowing how to use these heterogeneous resources – Originally constructed for various reasons, often with little attention to standards or interoperability – Have to pass data sets from one to another – Some involve user interaction Role of metadata Metadata is needed to enable discovery of resources and to indicate how they are to be used. • • • • Properties to help locate appropriate resources Check interoperability, suggest transformations Provenance of data sets Log of work-flows executed What is biodiversity informatics? • The preceding project, among others, shows that the challenges facing biodiversity informatics include not only – Describing the diversity of life at all levels of organisation, so that biologists can understand, conserve and exploit it, • But also – Inventing ways to describe the ever-increasing diversity of information resources and analysis tools available, so that users can find and use them A challenge to link resources • It is potentially very difficult to link all these resources together • Much attention is currently being given to: – Providing unique identifiers for data objects – Which can return metadata about themselves – Which can be stitched together into a distributed collaborative information system: see the biodiversity informatics organisations TDWG and GBIF (later) End