9 Genomic Techniques and How to Apply Them to Marine Questions
339
et al. 2008). It acts as a central repository for annotated eukaryotic genomes, ranging from Saccharomyces cerevisiae to human. The web interface allows the genome
to be browsed on several different levels, starting at the chromosomes and ending at
single bases. It also includes a sophisticated data mining mechanism.
The systems described here are only examples. Many more systems exist, with
new and improved systems being published every year. Almost every large sequencing project also offers its own web site with background information and specialized
databases.
9.3.3.3 Protein Annotation Tools
A protein function is often associated with the presence of a particular signature
or motif. Understanding the function of a protein is also often correlated with the
determination of its subcellular location. Consequently it is better to use multiple
methods and information resources for protein sequence annotation than a single
sequence similarity search tool such as those presented in the previous section.
In the first part, we will focus on InterPro, an integrated database for protein
function and structure analysis and InterProScan, a search tool, which combines the
protein signature recognition methods from the InterPro database and allows the
user to query an unknown sequence against the database in one step in order to aide
prediction of protein function.
In the second part, we will present two reliable prediction programs used for the
determination of the subcellular location: TMHMM for the detection of membranespanning segments and topology, and SignalP for the detection of signal peptides.
(a) InterPro and InterProScan
InterPro (Mulder et al. 2007) stands for Integrated Resource of Protein Families,
Domains and Functional Sites. Proteins or protein domains belonging to a particular
family usually share conserved regions often in correlation with evolution. Some
are directly important for function and some play a role in the preservation of the
protein 3D structure. The identification of such areas of sequence similarity allows
the determination of a unique signature for a particular protein family or domain.
The different proteins belonging to a protein family can then be distinguished from
other proteins from another family by their signature.
The InterPro consortium is currently composed of eleven member databases
located in Europe and the USA: UniProt (see also Section 9.3.5.2), PROSITE, Pfam,
PRINTS, ProDom, SMART, TIGRFAMs, PIRSF, SUPERFAMILY, Gene3D, and
PANTHER. The member databases have been constructed with different internal
representations such as profiles or position specific scoring schemes (Table 9.4).
Apart from protein signatures from the ProDom database, which are automatically
generated from the UniProt sequence database, the protein signatures from the other
member databases are manually curated. Refer to the member database homepages
for further details on the methodology and criteria they utilize.
By combining several protein signature databases with different strengths
and weaknesses, InterPro provides an integrated tool for protein function and
Précédent

- 350/410

Suivant