GeneFisher-P - User-Level Workflow Design: A Bioinformatics Perspective - page 88

then straightforward (left side of the figure): s is run through and at each po-
sition the theoretically possible primers (from minimum to maximum primer
length) are tested against the given global and 3' end properties. If a sub-
sequence of s passes all tests, it is added to the pool of possible (forward)
primers. Then, s is reversed and complemented and the procedure is repeated
in order to obtain a pool of primers for the reverse direction as well. Finally
(right side of the figure), primer pairs are built by matching each primer from
the forward pool to each primer from the backward pool. Only those pairs
are kept that conform to the given criteria regarding the size of the amplified
region, the avoidance of dimer structures (such as hairpin loop structures and
primer-primer interactions) and melting temperature compatibility.
Fig. 4.2 Primer design for a sequence s
As mentioned in the previous section, the PCR method can also be used
for detecting particular DNA segments in mixtures of DNA from various
organisms. Suitable degenerate primers can be designed from sequences of
homologous genes, based on the assumption that genes with related function
from different organisms have a high sequence similarity. In this case, the
primer design process starts with the computation of a multiple sequence
alignment from the homologous sequences. A consensus sequence is then cal-
culated from the alignment and used for the primer design as described above.
4.1.3 GeneFisher
The GeneFisher software [109] implements the method for designing degener-
ate primers described above. The term “gene fishing” refers to the technique
of using PCR to isolate a postulated but unknown target sequence from a
pool of DNA.
GeneFisher accepts single or multiple DNA and protein sequences as in-
put. As primers are always calculated for a single DNA sequence, multiple
input sequences are aligned using alignment programs such as ClustalW (cf.
Section 3.1.3), DCA [312] or DIALIGN [226]. From the alignment, a consen-
sus sequence is derived and used as input for the primer calculation step. If