Information Technology Reference
In-Depth Information
Considering all solutions until a search length of 6, the synthesis finds 650
possible workflows with the available services in the completely unconstrained
case. Applying the domain constraints reduces the number of returned solu-
tions to an easily accessible set of 8 solutions. Additional constraints that en-
force the use of a
DNA backtranslation
and a
Sequence alignment
service (as
required for an input consisting of multiple amino acid sequences) even lead to
one single solution. In contrast, when avoiding the use of
DNA backtranslation
and
Sequence alignment
services (as required for single amino acid sequences),
the specification is over-constrained and no solution is found, that is, the min-
imal GeneFisher-P workflow that directly calls the
GeneFisher
service with
the input sequence is not recognized by the synthesis. The reason is that, ac-
cording to the domain model,
GeneFisher
requires a
Nucleic acid sequence
as input, while
ReadSequence
produces a more general
Sequence
.Inorderto
fix this problem, additional services with more concise interface descriptions
would have to be introduced, for instance separate
ReadNucleotideSequence
and
ReadAminoAcidSequence
services that allow for a more fine-grained con-
sideration of the involved data types.
Search WWH ::
Custom Search