Semantic resources project/Antibodies/Discussions with Don Hatfield

Notes from Elizabeth's meeting with Don Hatfield (Alzforum Antibody Database Curators) 20 May 2009

On annotations


 * In addition to annotate specific antibodies to an individual research statement, antibodies should also be annotated as a group to a published paper, a hypothesis, or an experimental protocol.
 * On the Alzforum paper database, some papers are annotated with materials used already, Elizabeth will generate a list of pmids from SWAN to see if we could compare and find annotations already done in Alzforum papers
 * Many antibody entries in Alzforum already linked to published papers

On Sources of databases

Should also look at Millipore and Santa Cruz if we are considering commercial vendors as the data sources

On Searching

Alzforum antibody records reflects manufacturer's descriptions - and there is a taxonomy of reactivity species that was built based on the data sources - should review how to map or use that taxonomy. Don noted that sometimes the antibodies work against a specific transgenic animal, and the manufacturer does not disclose that info on their datasheet.
 * ability to search/visualize available antibodies based on positions in the peptides would be very useful
 * desepately seeking antibody section in Alzforum has a rich archives of questions from researchers - a great source of reviewing how to build the search criteria

Others

The Alzforum company database stored info re: the manufacturers of antibodies (Don noted that this is probably one of the most comprehensive database of antibody providers)

Immunogen and Epitope Notes from Don

Usually not covered in the description of immunogen and epitope is the immunization process itself. If you want a good description of this process, look at the data-sheets in the Developmental Studies Hybridoma Bank at the University of Iowa, http://dshb.biology.uiowa.edu/. For instance look at the data sheet for this antibody, which has much more information on the immunization than does the average data sheet: open the page http://dshb.biology.uiowa.edu/Antibody-list/crystal-protein and pick a "product form" and then click the data sheet row where it says "click here". I do not believe the ARF Antibody database needs to contain such information, but the SWAN version may want to. Most data sheets do not have much about the process.

The information about the immunogen (more specific than but still a form of the Antigen) includes the source of the immunogen protein (what kind of animal? was it natural or synthetic? was it recombinant? prokarotic recombinant? a mix of proteins? proteins from a particular organ of a particular organism? a particular protein or mix of proteins from a particular organ of a particular organism? or from a cell line?), the amino-acid span of the immunogen (especially when different from the epitope, see details below),  the conjugation (is it conjucated? with what? which end? residue(s) added?)

Note: this description of amino-acid spans applies both to the immunogen and the epitope. The epitope is the place where the antibody, made by the host animal as a response to the immunogen, will attach itself to the antigen protein. In many cases, but not all cases, the on-off amino acid sequence(s) of the immunogen define the epitope as well, and no separate specification of the epitope needs to be given. But often the amino acid sequence(s) of the epitope is/are more restrictive. The description of these sequences is usually of two alternative types: either a start-stop range or as one or more surrounds of an amino acid in a special state (typically phosphorylized). A start-stop range may be defined as two numbers, or by a region name such as N-terminal, C-terminal, C-terminal-fragments/CTF, full-length, or 'gleep' region, where 'gleep' might be "active" or "extracellular" or "internal" or "nuclear binding factor" (I know of no extensive list of region names in use, though there may be one, and it will surely grow. The 'surround' approach to description usually does not specify the size of the surround, but lists the amino acids of interest, along with their positions in the sequence. It is not assumed that everything between the first and last interesting amino acid is part of the epitope, though often there are only two amino-acids, and they are close together.  The other part of the description is the reason for the interest, typically phosphorylation.