predict.cancergram_model {CancerGram} | R Documentation |
Predict anticancer peptides
Description
Recognizes anticancer peptides using the CancerGram algorithm.
Usage
## S3 method for class 'cancergram_model'
predict(object, newdata, ...)
Arguments
object |
|
newdata |
|
... |
further arguments passed to or from other methods. |
Details
CancerGram requires the external package, CancerGramModel, which
contains models necessary to perform the prediction. The model
can be installed using install_CancerGramModel
.
Predictions for each protein are stored in objects of class
single_cancergram_pred
. It consists of three elements:
- seq
Character vector of amino acid sequence of an analyzed peptide/protein
- all_mers_pred
Matrix of predictions for each 5-mer (subsequence of 5 amino acids) of a sequence. Each row corresponds to one mer and columns to predicted classes (ACP, AMP or negative). Prediction value indicates probability that a 5-mer possesses anticancer activity (acp), antimicrobial activity (amp) or none of them (neg).
- single_prot_pred
One row matrix of a single prediction value for a whole peptide/protein. Its value corresponds to the probability that a peptide/protein exhibits anticancer activity, antimicrobial activity or none of them.
Value
list
of objects of class single_cancergram_pred
.
Each object of this class contains analyzed sequence, values of predictions
for 5-mers and result of the prediction for the whole peptide/protein.