|
ePitope Informatics is an Internet-based, applied bioinformatics
company specialising in protein annotation. A particular focus is epitope prediction for the identification and targeting of protein B cell epitopes.
Protein services at a glance
|
Consensus epitope prediction
(B cell) |
- Antigenicity (3
algorithms)
- Flexibility (1
algorithm)
- Hydrophilicity
(2 algorithms)
- Hydrophobicity
(5 algorithms)
- 2° structure prediction (2 algorithms)
- Beta-turn structure analysis
- PROSITE pattern
motif searching
- Rank of
epitopes as likely antigens
|
|
|
|
|
|
|
Epitope further analysis |
- BLASTP database
searching
(optimised for epitope queries)
- Protease
cut-site mapping
|
|
|
|
|
|
|
Epitope mapping to target protein solved 3-D structure |
- Degree of
epitope surface exposure
- Association
with beta-turn structures
- Rank of
epitopes as likely antigens
|
|
|
Epitope prediction
Our initial approach to epitope prediction (B cell) includes use of the
following algorithms (an easy-to-follow explanation of each algorithm
is provided with commercial service e-Reports):
|
 Protein antigenicity prediction algorithms
 |
Parker |
|
|
Protrusion Index (Thornton) |
|
|
Hopp and Woods |
|
|
Protein hydrophobicity algorithms
|
|
Kyte and Doolittle |
|
|
Janin |
|
|
Sweet and Eisenberg |
|
|
Manavalan |
|
|
Fauchere |
|
|
Protein hydrophilicity algorithms
|
|
Goldman, Engelman and Steitz (GES) |
|
|
von Heijne |
|
|
Protein flexibility prediction algorithm
|
|
Karplus and Schulz |
|
|
Protein secondary structure prediction
algorithms
|
|
Garnier and Robson |
|
|
Chou and Fasman |
|
|
|
|
Related protein annotation and analysis
Our service includes:
|
Epitope mapping to solved
protein 3-D
structure (if available) - see below.
Identification of amino acid sequence pattern motifs (PROSITE).
Reviewing available
published literature and specialist protein
databases for further information relating to your target protein
and predicted epitopes.
Comment on N- and C-terminus regions.
The effect of varying
algorithm window size (where applicable).
Prediction of transmembrane spanning regions.
Signal/cell sorting sequence prediction.
Target protein amino acid composition.
Target protein
physicochemical properties: including molecular
mass, isoelectric point, and molar extinction coefficient (280 nm).
Epitope peptide synthesis recommendation.
Recommendations for conjugating epitope peptides to carrier
protein.
|
Consensus epitope prediction
From the above analysis, we generate a list of consensus predicted
epitopes, ranked in order of their likely antigenicity. This data is
included in your service e-Report (see below).
Consensus predicted epitopes show:
High values for predicted antigenicity, flexibility and surface
location.
A high degree of association with predicted β-turns (secondary
structures frequently associated with antigenic sites,
particularly when they are adjacent to β-sheets or
α-helical
structures, in regions of hydrophilicity and flexibility).
Low and high values, respectively, for predicted hydrophobicity
and hydrophilicity. However, not all antigenic sites are
hydrophilic, hence the value of applying the above algorithms
and a consensus approach to epitope prediction.
Currently, ePitope Informatics identifies and ranks consensus
epitopes on the basis of how strongly a particular sequence is
identified as a predicted epitope within a particular algorithm,
combined with the frequency with which this sequence is identified
as a potential epitope across all algorithms, and from other
approaches that we use.
As part of our on-going research into epitope prediction, we aim to
test the validity of this approach for identifying and ranking
consensus epitopes.
Importantly, therefore, until the outcome of this work is known, the
final decision regarding which predicted epitope(s) you choose for
use in your work, is yours. Our role is to provide you with enough
information to greatly assist you making this choice.
Your final choice of epitope(s) will also be aided, possibly even
greatly influenced by, further information about predicted epitopes
that ePitope Informatics can provide (see below).
Further epitope analysis
For each target protein we offer further analysis of up to four
consensus predicted epitope sequences that includes investigation of:
their identity to databased protein sequences
the presence of protease cleavage sites
the presence of amino acid sequence motifs
their location on the solved 3-D structure (if available) of the
target protein of interest.
By default, we select for further analysis, consensus predicted
epitopes that overlap any target protein subsequence regions of
interest identified by you. In this situation, further analysis can
provide:
information that can aid, possibly even greatly influence, which
epitopes you choose for use in future work
molecular biological characterisation of consensus predicted
epitopes and, therefore, the protein they belong to
information that can be exploited in research projects
downstream of antibody production.
More detail
Database searching
ePitope Informatics uses BLAST (Basic Local Alignment Search Tool)
to search composite protein sequence databases held at the National
Centre for Biotechnology Information (NCBI). These searches, which
are performed from within Accelrys Discovery Studio software on
a maximum of 4 consensus predicted epitopes, retrieve sequences
showing identity to consensus predicted epitopes.
Protease site mapping
We analyse consensus predicted epitopes and flanking sequences for
the presence of protease cleavage sites, using the PABASE database
of proteolytic agents (OMIGA™ 2.0), and additional proteolytic sites
in our own database. This analysis is performed on a maximum of 4
consensus predicted epitopes.
Protein sequence motif searching
We analyse consensus predicted epitopes and flanking sequences for
the presence of protein sequence motifs using the PROSITE
database, additional sequence motifs obtained from published literature relating to target molecules, and extra motifs in our own database. This analysis is performed from within Accelrys Discovery Studio software, on a maximum of 4 consensus predicted epitopes.
Analysis of target protein 3-D structure
Newly updated graphics workstation and stereoscopic 3-D viewing hardware ("Stereo3D"/CrystalEyes® 3), for use with Discovery Studio
software from Accelrys, allows our introduction of new molecular modelling services for antibody epitope prediction and mapping to solved protein 3-D structure.
New
service options will be added to our online
quotation request form as they become available.
Currently, we perform protein 3-D analysis using several molecular 3-D viewing programs including Discovery Studio software
from Accelrys. Modelling potentially antigenic regions in this manner increases the accuracy of epitope prediction.
We map predicted antibody epitopes to the solved 3-D structure of your target protein (if available), and report on their location in relation to other
protein structural and functional sites.
For up to six consensus predicted epitopes, we report on their
degree of surface location, surface exposure of component amino
acids, and the epitope's association with actual rather than
predicted β-turns; β-turns being secondary structures frequently
associated with antigenic sites, particularly when they are adjacent
to β-sheets or α-helical structures in regions of hydrophilicity and
polypeptide chain flexibility.
Our service report also provides high quality colour images of different 3-D views of your target molecule annotated with predicted epitopes. These
Figures show labelling of surface exposed epitope amino acids.
Such an analysis was used to construct the animated and
non-animated images showing predicted epitopes at the surface of
the human DNA repair protein, MGMT, shown on our
Home page (and elsewhere on this site).
Then we rank identified surface-located epitopes on the basis of
a combination of their:
 rank as a
consensus predicted epitope
degree of surface exposure
association with β-turns
Similarly to the identification and ranking of consensus predicted
epitopes, we aim to test the validity of this approach for ranking
identified surface-located epitopes, as part of our on-going
research into epitope prediction.
Importantly, therefore, until the outcome of this work is known, the
final decision regarding which predicted epitope(s) you choose for
use in your work, is yours. Our role is to provide you with enough
information to greatly assist you making this choice.
 e-Report
We send you an e-Report as part of all our commercial services.
For secure server orders, we send e-Reports (in Adobe PDF format)
via the Internet by
Tumbleweed Secure Messenger (IME) (free of
charge). For standard server orders, e-Report PDF files are delivered
by standard e-mail or
secure e-mail (click here for more details).
e-Reports are in Adobe PDF format. We use this format as it can be
accessed across nearly all computer operating system platforms (e.g.
Windows, Mac, and most versions of Unix). e-Reports are about 50
pages in length, and are enabled with Adobe Acrobat bookmarks (to
aid their navigation) and active hyperlinks. Please note that the on-
screen version of your e-Report does not yield as high a resolution as
the printed copy.
e-Reports set out the results obtained from the above analysis of
your target molecule of interest, in mixed text and graphics format.
This includes our annotation of the target molecule's primary
sequence with predicted epitope sequences and protein
secondary structure.
Consensus predicted epitopes subjected to further analysis (see
above) are annotated with protease cut-sites, amino acid sequence
motifs, and predicted secondary structure obtained from output
derived for the whole molecule. In addition, if 3-D structure analysis
has been performed, your e-Report will contain output showing
actual secondary structure of your target protein.
Further output provided can include nice-view-like lists of databased
protein alignments (usually 100 - 300 per epitope), and cut-sites for
the 120 proteases in our database. Due to their large page length,
this output is supplied as extra PDF files in addition to your e-Report
PDF file.
Target protein 3-D structure analysis provides a series of graphics
images of the solved 3-D structure of your molecule of interest,
annotated with up to 6 consensus predicted epitopes. This
output also shows surface-located, partially exposed, buried
residues, and β-turns associated with each epitope.
e-Reports also include an easily understood description and
explanation of each of the above algorithms, and any other
approaches that we use to analyse your sequence/structure of
interest. They also include easy to follow guidance for your
assessment and evaluation of the data we provide.
Your e-Report, which is optimally viewed using the latest version of
Adobe Acrobat Reader (version 4.x onwards required), initially is
disabled for printing, and for selecting text and graphics for use in
other applications. Once we have received payment from you for the
service we have performed, these functions will be enabled (see
"Payment" section).
Optional hardcopy report
In
 addition, we offer you the option of ordering a hardcopy version
of this report, printed in colour and black and white
on high quality, "bright white" paper, and loosely
bound in a plastic report folder. This is a printout
of the main report PDF file in the e-Report package
we send you.
This printed version of your report does not include hardcopy of the
content of extra PDF files in an e-Report package (see above).
In the instance of a hardcopy report alone being ordered (i.e. no e-
Report), these extra PDF files will be sent to you along with the
hardcopy report on 3½" disks or CD.
Hardcopy reports are courier delivered to you by FedEx or TNT. We
notify you by e-mail when reports have been despatched, and we
give you a Waybill or Consignment Number that allows you to track
delivery via the Web.
Future services
One of our goals is to remain relevant to scientists' needs by
constructively responding to feedback.
Therefore, we would appreciate any input you are willing to share regarding services we might provide in the future. Possible future service options include:
prediction of discontinuous (conformational) epitopes*
protein identity/homology modelling*
T cell epitope prediction and analysis
*Assistance now offered as part of a client-tailored service.
This page was last updated:
Tuesday January 22, 2008
Copyright © 2000 - 2008 ePitope Informatics
|
