signalHsmm - prediction of signal peptides

Share:

Description

Using hidden semi-Markov models as a probabilistic framework, signalHsmm is new, highly accurate signal peptide predictor for eukaryotic proteins.

Details

Secretory signal peptides are short (20-30 residues) N-terminal amino acid sequences tagging among others tag among others hormons, immune system proteins, structural proteins, and metabolic enzymes. They direct a protein to the endomembrane system and next to the extracellular localization. All signal peptides possess three distinct domains with variable length and characteristic amino acid composition. Despite their variability, signal peptides are universal enough to direct properly proteins in different secretory systems. For example, artifically introduced bacterial signal peptides can guide proteins in mammals and plants.

The development of signalHsmm was funded by National Science Center (2015/17/N/NZ2/01845).

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
few_predictions <- run_signalHsmm(benchmark_dat[1:3])
#see all predictions
pred2df(few_predictions)
#summary one prediction
summary(few_predictions[[1]])
#plot one prediction
plot(few_predictions[[1]])

#have fun with GUI
## Not run: 
gui_signalHsmm()

## End(Not run)