T or select a valid benchmark dataset to train and test the predictor; (ii) formulate the protein samples with an effective mathematical expression that can truly reflect their intrinsic correlation with the target to 
be predicted; (iii) introduce or develop a powerful algorithm (or IQ1 engine) to operate the prediction; (iv) properly perform cross-validation tests to objectively evaluate the anticipated accuracy of the predictor; (v) establish a user-friendly web-server for the predictor that is accessible to the public. Below, let us describe how to deal with these steps.correlation with the target to be predicted [34]. To realize this, the pseudo amino acid composition (PseAAC) was proposed [21] to replace the simple amino acid composition (AAC) for representing the sample of a protein. Ever since the concept of SIS3 site PseAAC was introduced in 2001 [21], it has penetrated into almost all the fields of protein attribute predictions, such as predicting protein submitochondrial localization [35], predicting protein structural class [36], predicting DNA-binding proteins [37], identifying bacterial virulent proteins [38], predicting metalloproteinase family [39], predicting protein folding rate [40], predicting GABA(A) receptor proteins [41], predicting protein supersecondary structure [42], identifying protein quaternary structural attribute [43], predicting cyclin proteins [44], classifying amino acids [45], predicting enzyme family class [46], identifying risk type of human papillomaviruses [47], and discriminating outer membrane proteins [48], among many others (see a long list of references cited in [49]). Because it has been widely used, recently a powerful software called PseAAC-Builder [49] was proposed for generating various special modes of PseAAC, in addition to the web-server 
PseAAC [50] established in 2008. According to a recent review [34], the general form of PseAAC for a protein P can be formulated as P ?y1 y2 ?yu ?yV T ??Materials and Methods 1. Benchmark DatasetThe benchmark dataset Bench used in this study was taken from Verma et al. [2]. The dataset can be formulated asBenchz[{??where z contains 252 secretory proteins of malaria parasite, { contains S non-secretory proteins of malaria parasite, and the 252 symbol represents the union in the set theory. The same benchmark dataset was also used by Zuo and Li [4]. For reader’s convenience, the sequences of the 252 secretory proteins in z and those in { are given in Supporting Information S1.where T is a transpose operator, while the subscript V is an integer and its value as well as the components y1 , y2 , … will depend on how to extract the desired information from the amino acid sequence of P. The form of Eq.2 can cover almost all the various modes of PseAAC. Particularly, it can be used to reflect much more essential core features deeply hidden in complicated protein sequences, such as those for the functional domain (FunD) information [51,52,53] (cf. Eqs.9?0 of [34]), gene ontology (GO) information [54,55] (cf. Eqs.11?2 of [34]), and sequence evolution information [3] (cf. Eqs.13?4 of [34]). In this study, we are to use a novel approach to define the V elements in Eq.2. As is well known, biology is a natural science with historic dimension. All biological species have developed starting out from a very limited number of ancestral species. It is true for protein sequence as well [56]. Their evolution involves changes of single residues, insertions and deletions of several re.T or select a valid benchmark dataset to train and test the predictor; (ii) formulate the protein samples with an effective mathematical expression that can truly reflect their intrinsic correlation with the target to be predicted; (iii) introduce or develop a powerful algorithm (or engine) to operate the prediction; (iv) properly perform cross-validation tests to objectively evaluate the anticipated accuracy of the predictor; (v) establish a user-friendly web-server for the predictor that is accessible to the public. Below, let us describe how to deal with these steps.correlation with the target to be predicted [34]. To realize this, the pseudo amino acid composition (PseAAC) was proposed [21] to replace the simple amino acid composition (AAC) for representing the sample of a protein. Ever since the concept of PseAAC was introduced in 2001 [21], it has penetrated into almost all the fields of protein attribute predictions, such as predicting protein submitochondrial localization [35], predicting protein structural class [36], predicting DNA-binding proteins [37], identifying bacterial virulent proteins [38], predicting metalloproteinase family [39], predicting protein folding rate [40], predicting GABA(A) receptor proteins [41], predicting protein supersecondary structure [42], identifying protein quaternary structural attribute [43], predicting cyclin proteins [44], classifying amino acids [45], predicting enzyme family class [46], identifying risk type of human papillomaviruses [47], and discriminating outer membrane proteins [48], among many others (see a long list of references cited in [49]). Because it has been widely used, recently a powerful software called PseAAC-Builder [49] was proposed for generating various special modes of PseAAC, in addition to the web-server PseAAC [50] established in 2008. According to a recent review [34], the general form of PseAAC for a protein P can be formulated as P ?y1 y2 ?yu ?yV T ??Materials and Methods 1. Benchmark DatasetThe benchmark dataset Bench used in this study was taken from Verma et al. [2]. The dataset can be formulated asBenchz[{??where z contains 252 secretory proteins of malaria parasite, { contains S non-secretory proteins of malaria parasite, and the 252 symbol represents the union in the set theory. The same benchmark dataset was also used by Zuo and Li [4]. For reader’s convenience, the sequences of the 252 secretory proteins in z and those in { are given in Supporting Information S1.where T is a transpose operator, while the subscript V is an integer and its value as well as the components y1 , y2 , … will depend on how to extract the desired information from the amino acid sequence of P. The form of Eq.2 can cover almost all the various modes of PseAAC. Particularly, it can be used to reflect much more essential core features deeply hidden in complicated protein sequences, such as those for the functional domain (FunD) information [51,52,53] (cf. Eqs.9?0 of [34]), gene ontology (GO) information [54,55] (cf. Eqs.11?2 of [34]), and sequence evolution information [3] (cf. Eqs.13?4 of [34]). In this study, we are to use a novel approach to define the V elements in Eq.2. As is well known, biology is a natural science with historic dimension. All biological species have developed starting out from a very limited number of ancestral species. It is true for protein sequence as well [56]. Their evolution involves changes of single residues, insertions and deletions of several re.
