The discovery of peptides possessing high biological activity is very challenging due to the enormous diversity for which only a minority have the desired properties. To lower cost and reduce the time to obtain promising peptides, machine learning approaches can greatly assist in the process and even partly replace expensive laboratory experiments by learning a predictor with existing data or with a smaller amount of data generation. Unfortunately, once the model is learned, selecting peptides having the greatest predicted bioactivity often requires a prohibitive amount of computational time. For this combinatorial problem, heuristics and stochastic optimization methods are not guaranteed to find adequate solutions. We focused on recent advances in kernel methods and machine learning to learn a predictive model with proven success. For this type of model, we propose an efficient algorithm based on graph theory, that is guaranteed to find the peptides for which the model predicts maximal bioactivity. We also present a second algorithm capable of sorting the peptides of maximal bioactivity. Extensive analyses demonstrate how these algorithms can be part of an iterative combinatorial chemistry procedure to speed up the discovery and the validation of peptide leads. Moreover, the proposed approach does not require the use of known ligands for the target protein since it can leverage recent multi-target machine learning predictors where ligands for similar targets can serve as initial training data. Finally, we validated the proposed approach in vitro with the discovery of new cationic antimicrobial peptides. Source code freely available at http://graal.ift.ulaval.ca/peptide-design/.

The use of cyclic peptides in one-bead-one-compound libraries is limited by difficulties in sequencing hit compounds. Lacking a free N-terminal amine, such peptides cannot be sequenced by the Edman degradation approach, and complex fragmentation patterns are obtained by tandem mass spectrometry. To overcome this problem, we designed an alternative approach introducing a methionine residue within the macrocycle and as a linker to allow simultaneous ring-opening and release from the resin upon treatment with cyanogen bromide. The methionine linker was inverted relative to the peptide chain to allow the synthesis of cyclic peptides anchored by a lysine side chain and to avoid the presence of two C-terminal homoserine lactones on the released linear peptides. After MALDI-TOF MS/MS analysis, the peptides released from a single bead were sequenced manually and with a de novo sequencing software. The strategy described herein is compatible with commonly used amino acids and allows sequencing of cyclic peptides in one-bead-one-compound libraries, thus reducing the need for encoding.

ABSTRACT:A novel dual ring-opening/cleavage strategy to determine the sequence of cyclic peptides from one-bead-onecompound libraries is described. The approach uses a photolabile residue within the macrocycle and as a linker to allow a simultaneous ring-opening and cleavage from the beads upon UV irradiation and provide linearized molecules. Cyclic peptides of 5 to 9 residues were synthesized and the generated linear peptides successfully sequenced by tandem mass spectrometry.Peptide macrocycles are useful tools in chemical biology and medicinal chemistry to study and modulate proteins functions. 1 With great potential as therapeutic agents, they have gained a lot of interest in drug discovery. 2 Compared to their linear counterparts, cyclic peptides are more resistant to proteases and their increased conformational rigidity lowers the entropic cost of binding, making them tighter-binding to a given macromolecule. 3 The great degree of molecular diversity and complexity that can be accessed by simple changes in their sequence has prompted the use of cyclic peptides in combinatorial chemistry. The one-bead-one-compound (OBOC) approach, in which each bead carries many copies of a unique compound, has become a powerful tool in the drug discovery process. 4 However, the use of cyclic peptides in combinatorial OBOC libraries has been limited by difficulties in sequencing hit compounds after the screening. Lacking a free N-terminal amine, Edman degradation sequencing cannot be used on cyclic peptides and complicated fragmentation patterns are obtained by tandem mass spectrometry (MS/MS). 5 In this regard, a one-bead-two-compound approach on topologically segregated bilayer beads has been developed. 6 Initially introduced with an enzymatic shaving strategy 7 and later via chemical approaches, 8 topological bilayer segregation offers the opportunity to synthesize two compounds per bead, namely one which is exposed on the bead surface for screening (cyclic peptide) and the other found inside as a tag for sequencing and compound identification (linear peptide). More recently, Lim and co-workers reported a ring-opening strategy on cyclic peptoids to eliminate the need for encoding. 9 The approach involved the introduction of a cleavable alkylthioaryl bridge in the cycle to allow linearization of the molecule after the screening by thioether oxidation followed by nucleophilic displacement of the sulfone to generate a linear peptoid which can be sequenced by MS/MS. An attractive advantage of the ring-opening strategy over encoding methods is the absence of interference by the coding tag during screening since the same molecules are displayed inside and on the surface of the beads. Based on this strategy, we and other groups placed a methionine or a thioether bridge in cyclic peptides or peptoids and as a linker to allow a simultaneous linearization and compound release from the bead upon treatment with cyanogen bromide. 10 A similar approach was applied to cyclic depsipeptides where an aminolysis with NH 3 or hydrolysi...

A new methodology to couple peptide fragments on solid support using a traceless isocyanide-based multicomponent reaction is described. The approach uses a microwave-assisted on-resin Ugi four-component reaction to attach a carboxyl free peptide to a supported peptide bearing a free N-terminal amine via the formation of an N-protected amide bond at the ligation site. Afterward, the generated backbone amide protecting group can be efficiently removed by microwave-assisted acidolysis with trifluoroacetic acid to afford a fully deprotected peptide. This straightforward Ugi reaction/deprotection approach was applied to condense various fragment lengths and provided a variety of oligopeptides.