Insights into modifiers effects in differential mobility spectrometry: A data science approach for metabolomics and peptidomics

Utilizing a data-driven approach, this study investigates modifier effects on compensation voltage in differential mobility spectrometry–mass spectrometry (DMS-MS) for metabolites and peptides. Our analysis uncovers specific factors causing signal suppression in small molecules and pinpoints both signal suppression mechanisms and the analytes involved. In peptides, machine learning models discern a relationship between molecular weight, topological polar surface area, peptide charge, and proton transfer-induced signal suppression. The models exhibit robust performance, offering valuable insights for the application of DMS to metabolites and tryptic peptides analysis by DMS-MS.     

基于深度学习的保留时间预测方法的研究进展及应用

在基于液相色谱-质谱联用的蛋白质组学研究中,肽段的保留时间作为有效区分不同肽段的特征参数,可以根据肽段自身的序列等信息对其进行预测.使用预测得到的保留时间辅助质谱数据鉴定肽段序列可以提高鉴定的准确性,因此对保留时间预测的工作一直受到领域内的广泛关注.传统的保留时间预测方法通常是根据氨基酸序列计算肽段的理化性质,进而计算...     

Who Binds Better? Let Alphafold2 Decide!

Deep learning is revolutionizing structural biology to an unprecedented extent. Spearheaded by DeepMind's Alphafold2, structural models of high quality can be generated, and are now available for most known proteins and many protein interactions. The next challenge will be to leverage this rich structural corpus to learn about binding: which protein can contact which partner(s), and at what affinity? In a recent study, Chang and Perez have presented an elegant approach towards this challenging goal for interactions that involve a short peptide binding to its receptor. The basic idea is straightforward: given a receptor that binds to two peptides, if the receptor sequence is presented with both peptides together at the same time, AlphaFold2 should model the tighter binding peptide into the binding site, while excluding the second. A simple idea that works!     

Homology-free prediction of functional class of proteins and peptides by support vector machines

Protein and peptide sequences contain clues for functional prediction. A challenge is to predict sequences that show low or no homology to proteins or peptides of known function. A machine learning method, support vector machines (SVM), has recently been explored for predicting functional class of proteins and peptides from sequence-derived properties irrespective of sequence similarity, which has shown impressive performance for predicting a wide range of protein and peptide classes including certain low- and non- homologous sequences. This method serves as a new and valuable addition to complement the extensively-used alignment-based, clustering-based, and structure-based functional prediction methods. This article evaluates the strategies, current progresses, reported prediction performances, available software tools, and underlying difficulties in using SVM for predicting the functional class of proteins and peptides.     

Multidimensional Design of Anticancer Peptides

The computer‐assisted design and optimization of peptides with selective cancer cell killing activity was achieved through merging the features of anticancer peptides, cell‐penetrating peptides, and tumor‐homing peptides. Machine‐learning classifiers identified candidate peptides that possess the predicted properties. Starting from a template amino acid sequence, peptide cytotoxicity against a range of cancer cell lines was systematically optimized while minimizing the effects on primary human endothelial cells. The computer‐generated sequences featured improved cancer‐cell penetration, induced cancer‐cell apoptosis, and were enabled a decrease in the cytotoxic concentration of co‐administered chemotherapeutic agents in vitro. This study demonstrates the potential of multidimensional machine‐learning methods for rapidly obtaining peptides with the desired cellular activities.     

A critical cross-validation of high throughput structural binding prediction methods for pMHC

T-cells recognize antigens via their T-cell receptors. The major histocompatibility complex (MHC) binds antigens in a specific way, transports them to the surface and presents the peptides to the TCR. Many in silico approaches have been developed to predict the binding characteristics of potential T-cell epitopes (peptides), with most of them being based solely on the amino acid sequence. We present a structural approach which provides insights into the spatial binding geometry. We combine different tools for side chain substitution (threading), energy minimization, as well as scoring methods for protein/peptide interfaces. The focus of this study is on high data throughput in combination with accurate results. These methods are not meant to predict the accurate binding free energy but to give a certain direction for the classification of peptides into peptides that are potential binders and peptides that definitely do not bind to a given MHC structure. In total we performed approximately 83,000 binding affinity prediction runs to evaluate interactions between peptides and MHCs, using different combinations of tools. Depending on the tools used, the prediction quality ranged from almost random to around 75% of accuracy for correctly predicting a peptide to be either a binder or a non-binder. The prediction quality strongly depends on all three evaluation steps, namely, the threading of the peptide, energy minimization and scoring.     

Branched peptides as therapeutics

The concept of 'magic bullet', initially ascribed to immunoglobulins by Paul Ehrlich at the beginning of the 20th century and strengthened by the hybridoma technology of Kohler and Milstein in the mid 70s, can nowadays be attributed to different target-specific molecules, such as peptides. This attribution is increasingly valid in light of the explosion of new technologies for peptide library construction and screening, not to mention improvements in peptide synthesis and conjugation and in-vivo peptide stability, which make peptide molecules specific bullets for targeting pathological markers and pathogens. Today, hundreds of peptides are being developed and dozens are in clinical trials for a variety of diseases, demonstrating that the general reluctance towards peptide drugs that existed a decade ago has now been overcome. In spite of this progress, the development of new peptide drugs has largely been limited by their short half-life. Branched peptides such as Multiple Antigen Peptides (MAPs) were invented in the 80s by Tam [Tam, J.P., (1998) Proc. Natl. Acad. Sci. USA, 85, 5409] and have been extensively tested to reproduce single epitopes to stimulate the immune system for new vaccine discovery. In our lab we discovered that MAP molecules acquire strong resistance to proteases and peptidases. This resistance renders MAPs very stable and thus suitable for drug development. Here we report our experience with several MAP molecules in different biotechnological applications ranging from antimicrobial and anti toxin peptides to peptides for tumor targeting.     

Biological selection of peptides for poly(l-lactide) substrates

Short peptides that recognize the alpha form of poly( l-lactide) (PLLA) crystalline films were identified from a phage-displayed peptide library. An enzyme-linked immunosorbent assay (ELISA) revealed that the apparent binding constants of the phage clones for the alpha form of PLLA were greater than those of the unselected phage library. The specificity index for the alpha form of PLLA referred to a structurally similar atactic poly(methyl methacrylate) (at-PMMA), supporting the alpha form of PLLA specific binding of the selected phage. Amino acid residues with proton-donor lateral groups and hydrophobic alkyl groups were relatively enriched in a sequence of heptapeptides on the specific phage clones, thereby suggesting the presence of hydrogen bonding as well as hydrophobic interactions between the alpha form of PLLA and the peptides. Surface plasmon resonance (SPR) analysis revealed that the binding constant of the freed c22 heptapeptide (Gln-Leu-Met-His-Asp-Tyr-Arg) for the alpha form of PLLA was greater than those for reference at-PMMA, amorphous PLLA, and the beta form of PLLA. It was found that c22 peptide can recognize slight differences in PLLA polymorphs such as a crystalline state and an arrangement of PLLA functional groups.     

Structure prediction of cyclic peptides by molecular dynamics + machine learning

Recent computational methods have made strides in discovering well-structured cyclic peptides that preferentially populate a single conformation. However, many successful cyclic-peptide therapeutics adopt multiple conformations in solution. In fact, the chameleonic properties of some cyclic peptides are likely responsible for their high cell membrane permeability. Thus, we require the ability to predict complete structural ensembles for cyclic peptides, including the majority of cyclic peptides that have broad structural ensembles, to significantly improve our ability to rationally design cyclic-peptide therapeutics. Here, we introduce the idea of using molecular dynamics simulation results to train machine learning models to enable efficient structure prediction for cyclic peptides. Using molecular dynamics simulation results for several hundred cyclic pentapeptides as the training datasets, we developed machine-learning models that can provide molecular dynamics simulation-quality predictions of structural ensembles for all the hundreds of thousands of sequences in the entire sequence space. The prediction for each individual cyclic peptide can be made using less than 1 second of computation time. Even for the most challenging classes of poorly structured cyclic peptides with broad conformational ensembles, our predictions were similar to those one would normally obtain only after running multiple days of explicit-solvent molecular dynamics simulations. The resulting method, termed StrEAMM (Structural Ensembles Achieved by Molecular Dynamics and Machine Learning), is the first technique capable of efficiently predicting complete structural ensembles of cyclic peptides without relying on additional molecular dynamics simulations, constituting a seven-order-of-magnitude improvement in speed while retaining the same accuracy as explicit-solvent simulations.

The StrEAMM method enables predicting the structural ensembles of cyclic peptides that adopt multiple conformations in solution.     

Peptide–Column Interactions and Their Influence on Back Exchange Rates in Hydrogen/Deuterium Exchange-MS

Hydrogen/deuterium exchange (HDX) methods generate useful information on protein structure and dynamics, ideally at the individual residue level. Most MS-based HDX methods involve a rapid proteolytic digestion followed by LC/MS analysis, with exchange kinetics monitored at the peptide level. Localizing specific sites of HDX is usually restricted to a resolution the size of the host peptide because gas-phase processes can scramble deuterium throughout the peptide. Subtractive methods may improve resolution, where deuterium levels of overlapping and nested peptides are used in a subtractive manner to localize exchange to smaller segments. In this study, we explore the underlying assumption of the subtractive method, namely, that the measured back exchange kinetics of a given residue is independent of its host peptide. Using a series of deuterated peptides, we show that secondary structure can be partially retained under quenched conditions, and that interactions between peptides and reversed-phase LC columns may both accelerate and decelerate residue HDX, depending upon peptide sequence and length. Secondary structure is induced through column interactions in peptides with a solution-phase propensity for structure, which has the effect of slowing HDX rates relative to predicted random coil values. Conversely, column interactions can orient random-coil peptide conformers to accelerate HDX, the degree to which correlates with peptide charge in solution, and which can be reversed by using stronger ion pairing reagents. The dependency of these effects on sequence and length suggest that subtractive methods for improving structural resolution in HDX-MS will not offer a straightforward solution for increasing exchange site resolution.

Highly specific affinities of short peptides against synthetic polymers

We investigated polymer-binding 7-mer peptides that recognize differences in the polymer stereoregularity of all-purpose poly(methyl methacrylate)s (PMMAs) with simple chemical structures. Quantitative surface plasmon resonance measurements detected association/dissociation processes of the peptides against PMMA film surfaces, followed by an estimation of kinetic parameters such as association/dissociation rate constants and affinity constants. Greater association and smaller dissociation constants of the peptides were observed against a target isotactic PMMA than the structurally similar reference syndiotactic PMMA, followed by greater affinity constants against the target. A c02 peptide composed of the Glu-Leu-Trp-Arg-Pro-Thr-Arg sequence showed the greatest affinity constant (2.8x10(5) M(-1)) for the target, which was 41-fold greater than that for the reference, thus demonstrating extremely high peptide specificities. The substitution of each amino acid of the c02 peptide to Ala (Ala scanning) clearly revealed the essential amino acids for the affinity constants; the essential order was Pro5>Thr6>Arg7>Glu1>Arg4. In fact, the shorter 4-mer peptide composed of the C-terminal Arg-Pro-Thr-Arg sequence of the c02 peptide still demonstrated strong target specificity, although the N-terminal 4-mer peptide Glu-Leu-Trp-Arg completely lost its specificity. The possible conformations modeled with Molecular Mechanics supported the significance of the Arg-Pro-Thr-Arg sequence. The thermodynamic parameters of the c02 peptide suggested an induced fit mechanism for the specific affinity. The present affinity analyses of polymer-recognizing peptides revealed significant and general information that was essential for potential applications in peptidyl nanomaterials.     

Cell-specific delivery of a chemotherapeutic to lung cancer cells

We report that lung cancer-targeting peptides isolated from a peptide library can be used to deliver an active chemotherapeutic in a cell-specific fashion. The peptides were removed from the context of the phage and placed on a pegylated tetrameric scaffold. The tetrameric peptides were shown to block uptake of their cognate phage. The tetrameric peptides were coupled to doxorubicin, and their cytotoxicity against a panel of different cell lines was tested. Our data demonstrate that these targeting peptides can deliver an active anticancer agent in a cell-specific fashion, resulting in an increase of the therapeutic index of the targeted drug compared to systemic delivery. The efficacy of the peptide conjugate correlates to the affinity of the targeting peptide for a particular cell line. As such, we have demonstrated that cell-specific targeted drugs can be synthesized, even when the cell surface target is unknown.     

Peptide sequence prediction supported by correlation-associated networks in human cerebrospinal fluid

During the course of biosynthesis, processing and degradation of a peptide, many structurally related intermediate peptide products are generated. Human body fluids and tissues contain several thousand peptides that can be profiled by reversed-phase chromatography and subsequent MALDI-ToF-mass spectrometry. Correlation-Associated Peptide Networks (CAN) efficiently detect structural and biological relations of peptides, based on statistical analysis of peptide concentrations. We combined CAN with recognition of probable cleavage sites for peptidases and proteases in cerebrospinal fluid, resulting in a model able to predict the sequence of unknown peptides with high accuracy. On the basis of this approach, identification of peptide coordinates can be prioritized, and a rapid overview of the peptide content of a novel sample source can be obtained.     

Phage selection of bicyclic peptides binding Her2

Aberrant expression of the epidermal growth factor receptor Her2 has been implicated in various malignancies including breast cancer. Monoclonal antibodies and an antibody–drug conjugate targeting Her2 have found wide clinical application. Herein, we aimed at developing Her2-specifc ligands based on peptides that have a 100-fold smaller molecular weight than antibodies. Such peptides could potentially offer advantages in the development of ligand–drug conjugates, such as ease of synthesis and conjugation, higher molecule-per-mass ratios, and better tumor penetration. Panning of large bicyclic peptide phage display libraries against Her2 yielded a range of Her2-specific ligands having different formats and binding motifs. Strong sequence similarities among several of the isolated peptides indicated that they interact with Her2 in a specific manner. The best bicyclic peptide obtained after affinity maturation bound Her2 with a K_D of 304 nM. The diverse peptide ligands may offer valuable starting points for the development of high-affinity Her2 binders with potential application for tumor imaging and therapy.     

Machine learning techniques for the prediction of the peptide mobility in capillary zone electrophoresis

Three machine learning techniques including back propagation artificial neural network (BP-ANN), radial basis function artificial neural network (RBF-ANN) and support vector regression (SVR) were applied to predicting the peptide mobility in capillary zone electrophoresis through the development of quantitative structure-mobility relationship (QSMR) models. A data set containing 102 peptides with a large range of size, charge and hydrophobicity was used as a typical study. The optimal modeling parameters of the models were determined by grid-searching approach using 10-fold cross-validation. The predicted results were compared with that obtained by the multiple linear regression (MLR) method. The results showed that the relative standard errors (R.S.E.) of the developed models for the test set obtained by MLR, BP-ANN, RBF-ANN and SVR were 11.21%, 7.47%, 5.79% and 5.75%, respectively, while the R.S.E.s for the external validation set were 11.18%, 7.87%, 7.54% and 7.18%, respectively. The better generalization ability of the QSMR models developed by machine learning techniques over MLR was exactly presented. It was shown that the machine learning techniques were effective for developing the accurate and relaible QSMR models.     

Peptide structure prediction using distributed volunteer computing networks

Recent investigations to develop novel antimicrobial, antibiotical drugs have focused on the development of artificial protein peptides. As short peptides are naturally involved in many important biological processes in the cell and therefore target many kinds of cells. To functionalize peptides it is vital to design peptides, which can differentially target bacterial and eucariotic cells. Although the length of the peptides investigated in this study was limited to 16 amino acids, the number of possible peptide sequences is still too large to synthesize them in a trial- and error manner, therefore requiring a method for directed, but also high-througput peptide design. By predicting the structure of peptide proteins, this design process can be supported through structure-function analysis and peptide-membrane interaction simulation. In this investigation we could predict peptide structures de-novo, i.e. with the sequence information alone, using a massively parallel simulation scheme. We sample a sizable fraction of the peptide’s conformational space using Monte-Carlo simulations in the free-energy forcefield PFF02 on the volunteer computing network POEM@HOME. This forcefield models the protein’s native conformation as the global minimum of the free-energy. We could identify peptides of different topologies in a completely automated manner, which allows for the high-throughput screening of large peptide databases for their structural features, which would allow the rapid protopying of peptides needed for novel peptide design.     

ACPNet: A Deep Learning Network to Identify Anticancer Peptides by Hybrid Sequence Information

Cancer is one of the most dangerous threats to human health. One of the issues is drug resistance action, which leads to side effects after drug treatment. Numerous therapies have endeavored to relieve the drug resistance action. Recently, anticancer peptides could be a novel and promising anticancer candidate, which can inhibit tumor cell proliferation, migration, and suppress the formation of tumor blood vessels, with fewer side effects. However, it is costly, laborious and time consuming to identify anticancer peptides by biological experiments with a high throughput. Therefore, accurately identifying anti-cancer peptides becomes a key and indispensable step for anticancer peptides therapy. Although some existing computer methods have been developed to predict anticancer peptides, the accuracy still needs to be improved. Thus, in this study, we propose a deep learning-based model, called ACPNet, to distinguish anticancer peptides from non-anticancer peptides (non-ACPs). ACPNet employs three different types of peptide sequence information, peptide physicochemical properties and auto-encoding features linking the training process. ACPNet is a hybrid deep learning network, which fuses fully connected networks and recurrent neural networks. The comparison with other existing methods on ACPs82 datasets shows that ACPNet not only achieves the improvement of 1.2% Accuracy, 2.0% F1-score, and 7.2% Recall, but also gets balanced performance on the Matthews correlation coefficient. Meanwhile, ACPNet is verified on an independent dataset, with 20 proven anticancer peptides, and only one anticancer peptide is predicted as non-ACPs. The comparison and independent validation experiment indicate that ACPNet can accurately distinguish anticancer peptides from non-ACPs.     

Spontaneous membrane-translocating peptides by orthogonal high-throughput screening

Combinatorial peptide chemistry and orthogonal high-throughput screening were used to select peptides that spontaneously translocate across synthetic lipid bilayer membranes without permeabilization. A conserved sequence motif was identified that contains several cationic residues in conserved positions in an otherwise hydrophobic sequence. This 9-residue motif rapidly translocates across synthetic multibilayer vesicles and into cells while carrying a large polar dye as a "cargo" moiety. The extraordinary ability of this family of peptides to spontaneously translocate across bilayers without an energy source of any kind is distinctly different from the behavior of the well-known, highly cationic cell-penetrating peptides, such as the HIV tat peptide, which do not translocate across synthetic bilayers, and enter cells mostly by active endocytosis. Peptides that translocate spontaneously across membranes have the potential to transform the field of drug design by enabling the delivery of otherwise membrane-impermeant polar drugs into cells and tissues. Here we describe the chemical tools needed to rapidly identify spontaneous membrane translocating peptides.