See also
sinaps command
SINAPS paper.
SINAPS (Simple Non-Bayesian
Attribute Prediction Software) predict atrributes (traits) for a marker gene
sequence. From an abstract perspective, SINAPS is essentially the same algorithm
as SINTAX. A reference database is provided in
which each marker gene sequence is annotated with the trait to be predicted. A
random subset of words is extracted from the query sequence and used to find the
reference database sequence (top hit) with most words in common. This process is
repeated 100 times and the most frequently occurring top-hit trait is reported
as the predicted trait for the query. The number of iterations in which that
trait was the top hit is reported as the bootstrap confidence value.