HECTAR (heterokont subcellular targeting) has been designed to predict subcellular targeting for heterokont proteins. Heterokonts are supposed to have arisen by secondary endosymbiosis, this is the uptake of a red alga into a eukaryotic heterotroph. The endosymbiont was transformed into the heterokont chloroplast, showing a complex structure with four sourrounding membranes (contrary to two membranes for chloroplasts of land plants). Protein targeting into the chloroplast of heterokonts requires a N-terminal bipartite target peptide which consists of a leading signal peptide followed by a chloroplast transit peptide.
HECTAR is a prediction method which respects this complex chloroplast targeting. It is a modular method which consists of three decision modules. In each of these modules public available subcellular localisation methods are run and their outputs are combined using Support Vector Machines to predict the occurrence of one specific target peptide at a time. Alltogether five categories of subcellular targeting can be predicted by HECTAR: signal peptides, type II signal anchors, chloroplast transit peptides, mitochondrial transit peptides and those proteins which do not possess any of these N-terminal target sequences (marked as ''other localisation'').

HECTAR^SEC is a variant of HECTAR which allows the prediction of signal peptides and type II signal anchors for any eukaryotic protein.

The variant HECTAR^METFUN is dedicated to detect the presence of signal peptides, type II signal anchors and mitochondrial N-terminal target peptides for metazoan and fungal nuclear-encoded proteins.