A list of one or more UniProt primary accession numbers.
A 4-tuple containing the score results for a single pair of proteins.
The query protein.
The reference protein. Note: for predictors which produce a
symmetrical measure of similarity or relatedness between two
proteins, we would expect to see the same score and p-value
regardless of which way round the query and reference proteins
are.
A numeric score produced by the predictor. FuncNet does not
enforce any constraints on the range or distribution of scores
produced by its predictors, so these scores are not comparable
between different predictors.
A p-value representing the probability that any two randomly-
chosen proteins in this proteome would be given a rawScore
equal to or higher than the rawScore for this pair, by this
predictor. These probabilities are used by the integrator to
calculate an overall score between each protein pair from the
output of all the predictors.
Input type for the ScorePairwiseRelations operation.
A list of query proteins (as UniProt primary accessions)
with unknown or partially-known function.
A list of reference proteins (as UniProt primary accessions)
which are known to be involved in a particular biological
phenomenon, structure, process etc.
Output type for the ScorePairwiseRelations operation.
A list of score profiles for pairs of proteins which have
been assigned a score by the predictor. The maximum number of
elements in this list is QxR where Q is the number of query
proteins and R is the number of reference proteins. The list
may be shorter as the predictor is not guaranteed to return
a score for every possible pair.
This method runs the prediction algorithm on two sets of proteins (identified
by UniProt primary accessions), one query set and one reference set. The
predictor performs a pairwise comparison between each query-reference pair.