Drug discovery process plays an important role in identifying new investigational drug-likes and developing new potential inhibitors related to a determinate target, in biopharmaceutical field [1]. An alternative promising and efficient used to identify new active substances is Pharmacophore modeling method.We defined a new computational strategy protocol characterized by the use of bioinformatics online tools and by the application of locally installed tools, for lead candidates generation-optimization able to reduce the cycle time and cost of this process and to promote the next steps of study [2].Hence, we have tried to apply this new computational procedure, in a more detailed screening, of small bioactive molecules, searching and identifying new candidates as “lead compounds”, potentially able to inhibit biological target AKT1 human protein and its related molecular mechanisms [3].The workflow executed in our work has been characterized by a multi-step design, which concerns different topics: search in PDB database of a model structure for AKT1, pharmacophore modeling and virtual computational screening, biological evaluation divided in two parts (molecular validation of selected compounds and study of physical-chemical properties related to pharmacokinetic/pharmacodynamics prediction models). All these step have been performed through PHARMIT (http://pharmit.csb.pitt.edu) and Discovery Studio 4.5 platform.We selected the PDB structure 3O96 as the reference complex (protein-ligand), and we analyzed it by means of PHARMIT and Discovery Studio, to generate four different “pharmacophore models” with four different list of natural compounds.It is performed a thorough screening of compounds applying several filters, to find some good candidates as possible natural AKT1 allosteric inhibitors.The compounds that match a well-defined pharmacophore have been analyzed through direct molecular docking, for selecting only the best candidates and studying the protein-ligand interactions. Selected compounds have been investigated in more details, to trace their origin, by their chemical-physical properties.This information can help us to predict some plausible enzyme-catalyzed reaction pathways, through PathPred web-server and KEGG compound database, in order to highlight the most important reactions for biosynthesis of compounds and obtain PharmacoKinetics/PharmacoDynamics (PK/PD) models, to investigate the ADMET properties of these lead compounds and to study their behavior in some biological systems, for the next experimental assays.This new computational strategy has been very efficient in showing what could be good “lead compounds” and potential natural inhibitors of AKT1 and PI3K/AKT1 signaling cascade. Therefore, the next steps could be the experimental analysis of pharmacokinetics-pharmacodynamics and toxicity properties “in vitro/in vivo”, in order to evaluate the results obtained “in silico”.