The Bioinformatic-Harvester is a bioinformatic meta search engine at KIT Karlsruhe Institute of Technology for genes and protein-associated information. Harvester currently works for human, mouse, rat, zebrafish, drosophila and arabidopsis thaliana based information. Harvester cross-links >28 popular bioinformatic resources and allows cross searches. A ranking system similar to Google pagerank sorts the search results and displays the more relevant information. Harvester serves 10.000s of pages every day to scientists and physicians.
How Harvester works
Harvester collects information from protein and gene databases along with information from so called "prediction servers." Prediction server e.g provide online sequence analysis for a single protein. Harvesters search index is based on the IPI and UniProt protein information collection. The collections consists of:
- ~68.000 human, ~53.000 mouse, ~42.000 rat, ~51.000 zebrafish, ~35.000 arabidopsis and ~33.000 drosophila protein pages, which are curated and updated on a regular basis.
Harvester collects several types of information
...from the following databases:
- UniProt, world largest protein database
- SOURCE, convenient gene information overview
- Simple Modular Architecture Research Tool (SMART),
- SOSUI, predicts transmembrane domains
- PSORT, predicts protein localisation
- Homologene, compares proteins from different species
- gfp-cdna, protein localisation with fluorescence microscopy
- International Protein Index (IPI).
...are not collected, but crosslinked via iframes. Iframes are transparent windows within a HTML pages. The iframe windows allows up-to-date viewing of the "iframed," linked databases. Several such iframes are combined on a Harvester protein page. This method allows convenient comparison of information from several databases.
- NCBI-BLAST, an algorithm for comparing biological sequences NCBI.
- Ensembl, automatic gene annotation. EMBL-EBI and Sanger-Institute
- FlyBase is a database of model organism Drosophila melanogaster.
- GoPubMed is a knowledge-based search engine for biomedical texts.
- iHOP, information hyperlinked over proteins via gene/protein synonyms
- Mendelian_Inheritance_in_Man project catalogues all the known diseases.
- RZPD, German resources Center for genome research in Berlin/Heidelberg.
- STRING, Search Tool for the Retrieval of Interacting Genes/Proteins EMBL.
- LOCATE subcellular localization database (mouse).
- Genome_browser, working draft assemblies for genomes UCSC
- PolyMeta, meta search engine for Google, Yahoo, MSN, Ask, Exalead, AllTheWeb, GigaBlast
What one can find
Harvester allows a combination of different search terms and single words.
- Gene-name: "golga3"
- Gene-alias: "ADAP-S ADAS ADHAPS ADPS" (one gene name is sufficient)
- Gene-Ontologies: "Enzyme linked receptor protein signaling pathway"
- Unigene-Cluster: "Hs.449360"
- Go-annotation: "intra-Golgi transport"
- Molecular function: "protein kinase binding"
- Protein: "Q9NPD3"
- Protein domain: "SH2 sar"
- Protein Localisation: "endoplasmic reticulum"
- Chromosome: "2q31"
- Disease relevant: use the word "diseaselink"
- Combinations: "golgi diseaselink" (finds all golgi proteins associated with a disease)
- mRNA: "AL136897"
- Word: "Cancer"
- Comment: "highly expressed in heart"
- Author: "Merkel, Schmidt"
- Publication or project: "cDNA sequencing project"
- European Bioinformatics Institute
- Human Protein Reference Database
- Sequence profiling tool
- Liebel,U., & Kindler,B.,Pepperkok,R. (2004) 'Harvester': a fast meta search engine of human protein resources. Bioinformatics. 2004 Aug 12;20(12):1962-3. Epub 2004 Feb 26.
- Liebel,U., & Kindler,B.,Pepperkok,R. (2004) Bioinformatic "Harvester": a search engine for genome-wide human, mouse, and rat protein resources. Methods Enzymol. 2005;404:19-26