324
case of draft genome sequences, where often little is known about the function and
properties of the individual genes, or even when only an automatic annotation is given,
for example only by sequence comparison to a domain database.
• The UCSC Genome Browser allows a detailed view of the human genome along with
details on the properties of a gene. Particularly detailed are mRNA, exons, gene length
information, etc. In addition, the UCSC Genome Browser systematically provides
comparisons with other genomes (mammals, vertebrates).
• Swiss-Prot/UniProt: Here proteins are described particularly precisely and accurately.
Originally (Swiss-Prot), all proteins were annotated and examined by hand by experts,
but even now all annotations are checked in detail here.
The question of which annotation is most suitable cannot be answered in a general way,
because this always depends on the biological question. Let us briefly illustrate this with
our example of glutathione reductase. If we compare the annotation of the three different
databases, we can see differences in the database ID for glutathione reductase, for example
in GenBank and in the UCSD Genome Browser X54507 or in UniProt Q03504, despite
the uniform name designations. Thus, when using different databases, it is advisable to
always check if you actually have the same gene/protein between them. However, when
looking closer at the databases, it also becomes clear that UCSC and GenBank focus on
genomic position, whereas UniProt focuses more on biological function and interaction
context. So if you are only interested in the sequence, e.g. protein, you will find this in
GenBank and UniProt, whereas if you want to find out about the genomic region of the
gene, e.g. antisense or neighbouring genes, you should rather use UCSC and GenBank
(graphically visible). If you are looking for functional domains or interaction partners, you
will find more information in the UniProt database. So you can see that databases are
partly structured differently and have different foci, so it is up to you which database is
best suited. But there is one thing you should always keep in mind: Comparing several
databases is advisable in any case, because this way you can be sure to have found the right
information.
Question 6.6
Here you can think for yourself. It is important that the database/server should contain
trustworthy data (that it is also traceable where the data comes from), the user interface
should be easy to use, understandable and clear, but also up to date. It is also advantageous
to avoid overloaded pages and rather focus on one topic area, but provide further links (but
make sure that the links are always up to date) for individual analyses (a nice example is
our DrumPID database, which focuses on drugtarget interactions).
Question 6.7
It is important to check at regular intervals whether the data is still up to date or to enter
new data, but also to see whether the methodology is still up to date or whether there are
better procedures. Furthermore, you should make sure that cross-links to other websites or
websites necessary for the operation of the database are up to date and functioning.
20 Solutions to the Exercises