Prediction of Antithrombin Activity and Bioavailability of a Synthetic Peptide Using a Decision Tree Algorithm.

Background: Thrombosis is a major cause of mortality from cardiovascular disease (CVD). The use of known antithrombin drugs is limited by the presence of side effects and complications. Peptides may be promising antithrombin agents.

Objective: A peptide having the amino acid sequence QLSNGLFLFVDYLWW, designated as QW-13, was designed, synthesized, and used as a research object to evaluate the efficiency of the algorithm.

Methods: The solid-phase Fmoc (SPPS) method, followed by purification by high-performance liquid chromatography (HPLC) was used for synthesis. The molecular weight distribution of the peptide was estimated by mass spectrometry. Peptide identification was performed using the MALDI-TOF MS Ultraflex method. Mass spectra were analyzed using the Mascot program. To confirm the efficiency of the algorithm, the antithrombin effect of the peptide was studied, which was evaluated by the activated partial thromboplastin time (APTT) of human citrate blood plasma coagulation.

Results: A decision tree algorithm was developed to predict antithrombin activity and bioavailability of oral peptides. The top-down inductions of decision tree algorithms were used to create the algorithm. The decision tree is pruned if there is a mismatch in the peptide under study. Algorithm criteria or descriptors include amino acid sequence, number of amino acids, molecular weight, clinical potency, body distribution and metabolism index, plasma clearance, and half-life. According to the algorithm, if the result of "Antithrombin Peptide" is positive, it is concluded that it is bioavailable and effective for clinical use. To validate and evaluate the efficiency of the algorithm, a peptide containing antithrombin amino acid sequences was synthesized. The algorithm found the peptide to be antithrombin and bioavailable. The results of the biological activity of the peptide were confirmed in an in vitro experiment.

Conclusion: Employing a decision tree method to assess the antithrombin activity and bioavailability of peptides can facilitate the development of effective oral peptides, hence minimizing the development time and the quantity of in vitro and in vivo studies required to validate efficacy.