**4. Optimization methods for CPP-mediated cancer therapy and diagnosis**

Over the last decade, a great attention has been assigned to the importance of CPP on drug transportation of bioactive molecules in various preclinical studies. In fact, novel computational basics have been made in order to develop knowledge on CPPs [39].

Previously, different researchers have developed a few in silico algorithm approaches for CPP prediction (CPPpred) and screening to facilitate throughput CPP-based research. The in silico screening/prediction methods aimed on the use of scales of chemical characteristic, such as z-descriptors [40, 41]. It is generally followed by experimental validation to make it reliable with less cost and timeconsuming approach. Later on, other CPP prediction applied neural network (NN) strategies were developed and consist on introducing an N-to-1 NN. The network proceeds by a sequence of 5 to 30 amino acids in length, as input, and gives a prediction of how probably each peptide is to be cell penetrating [42]. This CPPpred offers an advantage since it was developed with repetition-reduced training and test sets.

Over the years, the commitment therapeutic importance of CPPs motivated other teams to develop the first version of CPP database, i.e., CPPsite which supports broad information on the promising use of CPPs [43]. The CPPsite manually created database of 843 experimentally described CPPs. Each consulting gives us data of the peptide involving peptide sequence, peptide name, nature of peptide, origin, chirality, uptake efficiency, subcellular localization, etc. A deep area of userfriendly tools has been integrated in this database like analyzing and browsing tools. Moreover, they have introduced other informations concerning peptide sequences such as secondary/tertiary structure and physicochemical properties of peptides.

This database version was then developed and updated as a CPPsite 2.0 and holds 1855 entries, including 1012 recent new entries [44]. The renovated version contains further data concerning chemically modified CPPs used on the in vivo model. In addition to other informations on delivered cargoes by CPPs (proteins, molecules, nanoparticles, DNA, RNA, etc.), secondary and tertiary structures of natural and chemical CPPs (including CPP with D-amino acids) were also predicted in view of their important role in the functionality of CPPs and stored in the database. Numerous tools for information browse and analysis are combined in this database and considered as a useful resource since it is compatible for all users, including smartphone and tablet.

CPP prediction sites are a promising assist to the researchers to design cell penetrating peptide, as well as making different modification and to investigate their effect on cell penetration potency [45].
