In Silico Screening and Optimization of Cell-Penetrating Peptides Using Deep Learning Methods

Abstract

Cell-penetrating peptides (CPPs) have great potential to deliver bioactive agents into cells. Although there have been many recent advances in CPP-related research, it is still important to develop more efficient CPPs. The development of CPPs by in silico methods is a very useful addition to experimental methods, but in many cases it can lead to a large number of false-positive results. In this study, we developed a deep-learning-based CPP prediction method, AiCPP, to develop novel CPPs. AiCPP uses a large number of peptide sequences derived from human-reference proteins as a negative set to reduce false-positive predictions and adopts a method to learn small-length peptide sequence motifs that may have CPP tendencies. Using AiCPP, we found that short peptide sequences derived from amyloid precursor proteins are efficient new CPPs, and experimentally confirmed that these CPP sequences can be further optimized.

Virtual Screening of Peptide Libraries: The Search for Peptide-Based Therapeutics Using Computational Tools

“…Cell-penetrating peptides attract attention for intracellular delivery as they can work as carriers for different classes of molecules (e.g., small compounds, peptides, and oligonucleotides), thus overcoming their low bioavailability issue. In this context, computational tools to design CPPs can be very supportive [60,61]. The free web tool "CellPPD" (i.e., http://crdd.osdd.net/raghava/cellppd/) can be, for example, implemented to predict CPPs with elevated accuracy [60].…”

“…The approach behind "CellPPD" relies on the exploitation of different peptide features (e.g., the type of residues in each position, two-residue motives, and physicochemical properties) to develop support vector machine (SVM)-based models through which it is possible to discriminate between CPPs and non-CPPS [60]. The likelihood of false positives is a problem that can be encountered in CPP prediction, and the "AiCPP" approach seems to provide a possible solution [61]. In detail, this is a deep learning-based method employing a negative control group made up of a vast set of peptide sequences from human reference proteins [61].…”

“…The likelihood of false positives is a problem that can be encountered in CPP prediction, and the "AiCPP" approach seems to provide a possible solution [61]. In detail, this is a deep learning-based method employing a negative control group made up of a vast set of peptide sequences from human reference proteins [61].…”

Investigating molecular descriptors in cell-penetrating peptides prediction with deep learning: Employing N, O, and hydrophobicity according to the Eisenberg scale

“…Park et al (2023) [ 23 ] developed a method for predicting CPPs based on DL called AiCPP with the aim of avoiding false-positive results, incorporating the LSTM algorithm. This work contributes to research into predictions of new CPPs with the discovery that short peptide sequences derived from amyloid precursor proteins are more efficient in permeating the cell membrane, despite not achieving the objective of reducing false-positive results in relation to other algorithms, with 88.6% accuracy against more than 90% of this metric achieved by other algorithms [ 23 ].…”

Advances in Drug Design and Development for Human Therapeutics Using Artificial Intelligence-II

“…Nevertheless, the identification of specific regions within a protein or peptide sequence with CPP properties is traditionally a laborious and time-consuming process involving experimental verification and optimization. To expedite this process and make it more efficient, Park et al [2] have introduced a CPP prediction model which leverages deep learning-based natural language processing techniques, shedding light on the essential sequence patterns required to enhance the CPP characteristics. Ultimately, this breakthrough not only streamlines the design of novel CPPs but also enhances their cell-penetrating properties, offering promising enhancements in drug delivery strategies.…”