Information Technology Reference
In-Depth Information
4.2.4 Further Variations of the GeneFisher-P
Workflow
The interactive and the batch processing workflows are only two examples of
useful GeneFisher workflow realizations. The journal article on GeneFisher-
P [177], for instance, describes further workflow variations, similar to the
alignment workflow examples in Section 3.2 and also incorporating external
services: alternative ClustalW implementations or other multiple sequence
alignment methods are used (for instance via the EBI [250, 167, 118] or
DDBJ [165] Web Services), or the workflows are extended by automated
data retrieval, for instance by retrieving the input sequences from a remote
database. In a variation that resembles the very interactive nature of the
original GeneFisher web application closer than the workflow shown in Figure
4.5, also intermediate results are displayed to the user and may be accepted
or rejected. In case of rejection, the workflow leads the control-flow back to
a previous process step, typically to a point where the appropriate data or
parameters can be modified before the corresponding part of the workflow is
executed again.
At the beginning of this section it was stated that the objective behind
GeneFisher-P was to create a GeneFisher version that could easily be batch-
processed in order to design large numbers or primers, and where single com-
ponents could be easily exchanged by services that meet the specific needs
better. Both these features where in fact applied in a chelicerate phylogeny
study at the University of Hamburg [50], where several pairs of PCR primers
were designed using GeneFisher-P. The multiple alignment was carried out
using Mafft [150] instead of ClustalW, and the backtranslation and consensus
calculation via the legacy BatCons tool was replaced by a new Ruby script
that implements backtranslation and consensus calculation with some spe-
cific features that cope better with the characteristics of the data that were
used in the study.
4.3 Constraint-Driven Design of GeneFisher-P
Workflows
This section shows how loose programming can be applied for the devel-
opment of the GeneFisher-P workflows. As the primer design process is
performed by a set of rather common services (alignment, backtranslation,
consensus calculation, etc.) the EDAM terminology (cf. Section 3.3.1) can
also be used for the annotation of the services in the GeneFisher-P domain
model. However, the domain model for GeneFisher-P is much smaller than
the one for EMBOSS, and hence less constraints are necessary for keeping
the search and solutions within a feasible range.
Search WWH ::
Custom Search