Welcome to EpoDB

EpoDB stands for "erythropoiesis database." It is a tool for researchers studying red blood cells to obtain information on genes of interest. It is also a working project for database management system development, schema and database design for gene expression information, and computational analysis of gene sequences and expression data. Therefore, we have been working both on content and functionality of EpoDB as well as interface tools for annotation and access to the database.

A practical guide to the different web pages comprising the EpoDB web site is given in "Description of EpoDB pages." The intent and methodology for creation of EpoDB are described in "Purpose of EpoDB" and "Construction of EpoDB."

Description of EpoDB pages

The EpoDB server can be viewed with or without "frames" depending on the capability of the web browser in use. Both versions have the same content; the frame version has in addition a listing of query page links.

Icons and links:
• EpoDB icon: Top left of page: link that returns the home page
• CBIL icon: Top right of page: link to the Computational Biology and Informatics Laboratory (CBIL) home page.
• Controlled Vocabulary: Bottom of page: These are lists of names used in gene name and gene family name queries, and in annotating gene expression data.
• EpoDB group: Bottom of page: Email link to the EpoDB group (if you browser is set up to do email).
• Schema link: Bottom of page: Schemas used in annotating and implementing the EpoDB database.
• Erythropoiesis link: page describing red cell differentiation and pictures of human red blood cell precursors.
• What's new in version 2.1: list of new features and expanded queries introduced in version 2.1. Major changes that were introduced in version 2.0 are also listed.
• Background Information: link to this page.
• BLAST search: You can search EpoDB for homologous sequences to a sequence you submit using the BLAST programs.
• Keyword list: The keywords used to search GenBank and Swiss-prot are given along with the current number of entries.
• Queries: links to seven different query pages are given. These are described in detail below. An abbreviated listing of the same query pages is given on the left in the frame version.
• Related sites: links are provided to:
  • The globin gene server, a database of selected mammalian beta-globin gene clusters containing annotated sequence and related experiments;
  • Transfac, a transcription factor database;
  • TRRD, a database of gene regulation elements by our collaborators at the Institute of Cytology and Genetics SB Russian Academy of Sciences;
  • NCBI, the Entrez nucleotide query page for GenBank;

Query pages:
• Find entries by gene name: After choosing a gene name from the menu list, a table is returned with all the EpoDB entries of that name. The table lists organism and links to EpoDB feature descriptions derived from GenBank, transcription unit information, the GenBank Entrez nucleotide query for the entry, and (where available) links to protein features derived from Swiss-prot. In addition to the EpoDB features, the original GenBank features and (where available) Transfac-derived transcription factor binding sites are given. The protein features also provide links to Medline. A graphic view of the entry features and sequence can be obtained with Java-supported browsers. This feature is useful for determining precise sequence locations.
• Find entries by gene family name: This query is similar to the "Find entries by gene name" query above except that a list of gene family (and protein function) names are given.
• Find entries by ID: This query provides a direct method for obtaining a description of gene features by entering either an EpoDB ID or a GenBank locus ID. Return of the actual entry sequence is optional.
• Find entries by text search: This query allows search of all EpoDB entries (either DNA or protein) using a Boolean text search. This query is useful for quickly browsing EpoDB.
• Find sequences by landmark: This query allows extraction of specified stretches of sequence from chosen EpoDB entries. Sequences for all entries of a given gene name or a given gene family (protein function) name can be retrieved simultaneously. The stretch of sequence desired is specified by giving a range relative to a specific gene landmark chosen from a menu. This query is useful for obtaining promoter sequences by specifying ranges of sequences upstream of the start of transcription. The retrieved sequences can be saved in FASTA format to a file or exported to TESS, a transcription element search software package which is part of CBIL.
• Find sequences by feature: This query is very similar to the "Find sequence by landmark query described above. Instead of specifying a range of sequence relative to a landmark, the user chooses from a menu of gene features. For example, the coding regions for all chosen genes can be obtained in this way.
• Find promoter box: This query uses GenLang, a pattern search program based on linguistics, to look for specific transcription element motifs in promoter sequences of specified genes. The user can enter any motif to be used for the search or choose from common erythroid promoter boxes.

The Purpose of EpoDB

The overall purpose of creating the EpoDB database is to provide a powerful tool for understanding control of gene expression during erythropoiesis. In particular, the database is being built to study the organization of transcription regulatory elements as well as the dynamics of transcription factor interaction with those regulatory elements. To accomplish this, the content of the database will include both structural information as well as gene expression information represented in novel schemas and integrated with data analysis tools and algorithms such as BLAST and GenLang . The sources of information are from GenBank, Swiss-Prot, Transfac/TRRD entries and the literature. The quality of data obtained from GenBank is improved by removing redundancy, and eliminating syntactic/semantic errors. More powerful queries of EpoDB are possible than of GenBank because of greater structure, consistent and uniform annotation, and controlled vocabularies. Collaborations will also be available that allow for secure data.

Construction of EpoDB

Currently, EpoDB contains approximately 3715 GenBank and 1500 Swiss-Prot entries . Schematic representation of the entries is evolving, however, a first pass schema has been completed. First pass controlled vocabularies for gene names and gene family names have also been constructed. Present capabilities are to extract features and subsequences (e.g. retrieve proximal promoter region, -500 to +20 around start of transcription, for all beta-globin genes); and find transcription factor motifs using TESS. Efforts are underway to remove redundancy of gene entries, to incorporate "virtual" sequences linking related but non-contiguous sequences, and to increase gene expression annotation. Future plans are to expand the queries available to take advantage of protein, gene regulation and gene expression information being entered into EpoDB. EpoDB focuses on gene expression during red cell development and serves as a model database that can be easily extended to other hematopoietic lineages.

Virtual Entries in EpoDB: ( Return)
Virtual entries are the products of merges between two or more Genbank entries. These are constructed to link fragments of a gene into a single entry and/or syntenic members of a gene family. The syntax for creating virtual entries is essentially the Genbank "join" feature with the addition of "gap"s to represent missing sequence of indeterminate length. Thus, a query for human alpha-spectrin will return a single entry consisting of 50 ordered exons instead of 50 entries containing individual exons. Likewise, a query for human alpha globin genes will return a single entry from zeta to theta (with intergenic gaps) instead of several entries including duplicates. A sample virtual entry for the goat alpha globin gene cluster is given below in prolog format.

class(entry_id(142526),yes,ref,virt,0,0,0).

na_seq_info(oid(142526),
            [locus_id([]),
             accession([]),
             na_seq_id(sid(142527)),
             length(10000),
             strand(text("ds")),
             mol(text("dna")),
             def(text("goat alpha-globin cluster")),
             extra_accessions([]),
             source([]),
             keywords([]),
             origin([]),
             date(std([year(1995),month(9),day(12)])),
             div(text("mam")),
             taxonomy(text("Eukaryota; Animalia; Metazoa; Chordata; Vertebrata; Mammalia; Theria; Eutheria; Artiodactyla; Ruminantia; Pecora; Bovidae.")),
	     organism(text("Capra hircus"))]).

comments(entry_id(142526),[text("contains the zeta, I alpha, and II alpha-globin genes")]).

na_seq(entry_id(142526),
       na_seq_id(sid(142527)),
       join([ za - gi(946):'1..568',
              g1 - gap,
              zb - gi(948):'1..1247',
              g2 - gap,
              a1 - gi(164123):'1..1894',
              g3 - gap,
              a2 - gi(164125):'1..1691' ]) ).

can_feature( source,         [location('za:1..a2:1691'),
                                can_feat_id(142554),
                                organism('Capra hircus'),
                                change(date(1995,9,27),author(cjs),[del(type(new)),add(type(source)),
                                    add(location('za:1..a2:1691')),add(organism('Capra hircus'))]
                                    )] ).

can_feature(can_feat_id(142540),
            entry_id(142526),
            type(cluster),
            location('za:1..a2:1691'),
            quals([gene_family_name([globin,'globin, alpha-like']),
                   components([can_feat_id(142528),can_feat_id(123846),can_feat_id(123858)])]),
            change([])).

can_feature(can_feat_id(142528),
            entry_id(142526),
            type(transcript_unit),
            location('za:1..zb:1247'),
            quals([gene_name(['zeta-globin']),
                   gene_family_name([globin,'globin, alpha-like']),
                   spatio_temporal_transcription([]),
                   components([can_feat_id(142529),can_feat_id(142530),can_feat_id(142531),can_feat_id(142532),
		               can_feat_id(142533),can_feat_id(142534),can_feat_id(142535),can_feat_id(142536),
                               can_feat_id(142537),can_feat_id(142538),can_feat_id(142539),
                   parents([])
                  ])]),
            change([])).

can_feature(can_feat_id(142529),
            entry_id(142526),
            type(mRNA_boundaries),
            location('za:216..zb:1020'),
            quals([]),
            change([])).

can_feature(can_feat_id(142530),
            entry_id(142526),
            type(mRNA),
            location(join('za:216..356','zb:538..742','zb:852..1020')),
            quals([]),
            change([])).

can_feature(can_feat_id(142531),
            entry_id(142526),
            type(exon),
            location('za:216..356'),
            quals([]),
            change([])). 

can_feature(can_feat_id(142533),
            entry_id(142526),
            type('CDS'),
            location('za:262..356,zb:366..570,zb:809..934'),
            quals([]),
            change([])).

can_feature(can_feat_id(142534),
            entry_id(142526),
            type('intron'),
            location('za:357..zb:365,'),
            quals([]),
            change([])).
   
can_feature(can_feat_id(142535),
            entry_id(142526),
            type(exon),
            location('zb:366..570'),
            quals([]),
            change([])). 

can_feature(can_feat_id(142536),
            entry_id(142526),
            type('intron'),
            location('zb:571..808'),
            quals([]),
            change([])).

can_feature(can_feat_id(142537),
            entry_id(142526),
            type(exon),
            location('zb:809..1020'),
            quals([]),
            change([])). 

can_feature(can_feat_id(142532),
            entry_id(142526),
            type('5\'UTR'),
            location('za:216..261'),
            quals([]),
            change([])).

can_feature(can_feat_id(142538),
            entry_id(142526),
            type('3\'UTR'),
            location('zb:935..1020'),
            quals([]),
            change([])).

can_feature(can_feat_id(142539),
            entry_id(142526),
            type(misc_feature),
            location('zb:1015..1020'),
            quals([feat_id(10794),gb_note('polyA signal')]),
            change([])).

The EpoDB group at CBIL consists of:

• Brian Brunk
  
• Juergen Haas
  
• Fidel Salas
  
• Chris Stoeckert
  
• and Chris Overton , our fearless leader.

Collaborators at the Institute of Cytology and Genetics, SB RAS, Novosibirsk, Russia in creating GERD:

• Nikolay Kolchanov
• Alexander Kel
• Olga Kel
• Olga Podkolodnaya
• Irina Stepanenko
• Elena Ananko
• Elena Ignatyeva
• Nikolay Podkolodny