Search

Abstract The amino acid sequences of peptides determine their self-assembling properties. Accurate prediction of peptidic hydrogel formation, however, remains a challenging task. This work describes an interactive approach involving the mutual information exchange between experiment and machine learning for robust prediction and design of (tetra)peptide hydrogels. We chemically synthesize more than 160 natural tetrapeptides and evaluate their hydrogel-forming ability, and then employ machine learning-experiment iterative loops to improve the accuracy of the gelation prediction. We construct a score function coupling the aggregation propensity, hydrophobicity, and gelation corrector C g , and generate an 8,000-sequence library, within which the success rate of predicting hydrogel formation reaches 87.1%. Notably, the de novo-designed peptide hydrogel selected from this work boosts the immune response of the receptor binding domain of SARS-CoV-2 in the mice model. Our approach taps into the potential of machine learning for predicting peptide hydrogelator and significantly expands the scope of natural peptide hydrogels.

Biosurfactants have garnered increased attention lately due to their superiority of their properties over fossil-derived counterparts. While the cost of production remains a significant hurdle to surpass synthetic surfactants, biosurfactants have been anticipated to gain a larger market share in the coming decades. Among these, glycolipids, a type of low-molecular-weight biosurfactant, stand out for their efficacy in reducing surface and interfacial tension, which made them highly sought-after for various surfactant-related applications. Glycolipids are composed of hydrophilic carbohydrate moieties linked to hydrophobic fatty acid chains through ester bonds that mainly include rhamnolipids, trehalose lipids, sophorolipids, and mannosylerythritol lipids. This review highlights the current landscape of glycolipids and covers specific glycolipid productivity and the diverse range of products found in the global market. Applications such as bioremediation, food processing, petroleum refining, biomedical uses, and increasing agriculture output have been discussed. Additionally, the latest advancements in production cost reduction for glycolipid and the challenges of utilizing second-generation feedstocks for sustainable production are also thoroughly examined. Overall, this review proposes a balance between environmental advantages, economic viability, and societal benefits through the optimized integration of secondary feedstocks in biosurfactant production.

Abstract Self‐assembling of peptides is essential for a variety of biological and medical applications. However, it is challenging to investigate the self‐assembling properties of peptides within the complete sequence space due to the enormous sequence quantities. Here, it is demonstrated that a transformer‐based deep learning model is effective in predicting the aggregation propensity (AP) of peptide systems, even for decapeptide and mixed‐pentapeptide systems with over 10 trillion sequence quantities. Based on the predicted AP values, not only the aggregation laws for designing self‐assembling peptides are derived, but the transferability relation among the APs of pentapeptides, decapeptides, and mixed pentapeptides is also revealed, leading to discoveries of self‐assembling peptides by concatenating or mixing, as consolidated by experiments. This deep learning approach enables speedy, accurate, and thorough search and design of self‐assembling peptides within the complete sequence space of oligopeptides, advancing peptide science by inspiring new biological and medical applications.

Abstract Supramolecular chirality plays an indispensable role in living and synthetic systems. However, the generation and control of filament chirality in the supramolecular hydrogel of short peptides remains challenging. In this work, as the first example, we report that the heterodimerization of the enantiomeric mixture controls the alignment, chirality, and stiffness of fibrous hydrogels formed by aromatic building blocks. The properties of the resulting racemic hydrogel could not be achieved by either pure enantiomer. Cryo-EM images indicate that the mixture of L and D enantiomers forms chiral nanofibers, the percentage of which can be readily controlled through stoichiometric co-assembly of heterochiral enantiomers. 2D NOESY NMR and diffusion-ordered NMR spectroscopy reveal that heterodimerization of enantiomers plays a crucial role in the formation of chiral nanofibers. Further mechanistic studies unravel the mechanism of supramolecular chirality formation in this two-component system. Molecular dynamics simulations confirm that the intermolecular hydrogen bond and π–π interaction of heterodimers play important roles in forming a chiral hydrogel. Furthermore, regulation of the adhesion and morphology of mammalian cells is achieved by tuning the relative ratio of L and D enantiomers at the same concentration. This work illustrates a novel strategy to control the supramolecular chirality of aromatic peptide hydrogels for materials science. Graphical Abstract

Active learning (AL) is an effective sample selection approach that annotates only a subset of the training data to address the challenge of data annotation, and deep learning (DL) is data-intensive and reliant on abundant training data. Deep active learning (DeepAL) benefits from the integration of AL and DL, offering an efficient solution that balances model performance and annotation costs. The importance of DeepAL has been increasingly recognized with the emergence of large foundation models that depend heavily on substantial computational resources and extensive training data. This survey endeavors to provide a comprehensive overview of DeepAL. Specifically, we first analyze and summarize various sample query strategies, data querying considerations, model training paradigms, and real-world applications of DeepAL. In addition, we discuss the challenges that arise in the era of foundation models and propose potential directions for future AL research. The survey aims to bridge a gap in the existing literature by organizing and summarizing current approaches, offering insights into DeepAL and highlighting the necessity of developing specialized DeepAL techniques tailored to foundation models. By critically examining the current state of DeepAL, this survey contributes to a more profound understanding of the field and serves as a guide for researchers and practitioners interested in DeepAL techniques.

Computing is a critical driving force in the development of human civilization. In recent years, we have witnessed the emergence of intelligent computing, a new computing paradigm that is reshaping traditional computing and promoting digital revolution in the era of big data, artificial intelligence, and internet of things with new computing theories, architectures, methods, systems, and applications. Intelligent computing has greatly broadened the scope of computing, extending it from traditional computing on data to increasingly diverse computing paradigms such as perceptual intelligence, cognitive intelligence, autonomous intelligence, and human–computer fusion intelligence. Intelligence and computing have undergone paths of different evolution and development for a long time but have become increasingly intertwined in recent years: Intelligent computing is not only intelligence oriented but also intelligence driven. Such cross-fertilization has prompted the emergence and rapid advancement of intelligent computing. Intelligent computing is still in its infancy, and an abundance of innovations in the theories, systems, and applications of intelligent computing is expected to occur soon. We present the first comprehensive survey of literature on intelligent computing, covering its theory fundamentals, the technological fusion of intelligence and computing, important applications, challenges, and future perspectives. We believe that this survey is highly timely and will provide a comprehensive reference and cast valuable insights into intelligent computing for academic and industrial researchers and practitioners.

The lack of effective treatments for pulmonary diseases poses a global health burden. The direct local gene therapy serves as one of the alternative administrations to treat pulmonary diseases. Compared with the conventional viral/nonviral system, the peptide–vector‐mediated in vivo lung gene therapies exhibit various benefits. However, the related clinical trials are still in their infancy. The major obstacle to the pulmonary delivery of gene cargoes may be the barriers from various extracellular mucosal layers and intracellular membranes. This review highlights the recent development of peptide‐based gene delivery systems and their applications. The peptide designing rules for the barrier‐permeable pulmonary gene delivery are described first. After briefly summarizing how oligopeptides facilitate the local gene therapy in lung tissue with the focus on cell‐penetrating peptides, the local delivery system of the polypeptide and several alternative hybridizing systems of the peptide with other types of materials are discussed. Finally, the blueprint and the remaining challenges in peptide designations are discussed before they enter into the real translation process.

The Longmaxi Formation in the southern Sichuan Basin is an important target for shale gas exploration and development. The characteristics and stages of structural development significantly impact shale gas preservation and enrichment. Taking the Longmaxi Formation in the Yanjin–Junlian area of the southern Sichuan Basin as an example and based on the results of surface and underground structural analysis, fluid inclusion tests, apatite fission track experiments, and burial-thermal evolution history analysis, a comprehensive study of the development characteristics and structural stages of the Longmaxi Formation was carried out, and an evolution model was developed. (1) The Longmaxi Formation of the Yanjin–Junlian area has been affected by multistage structural movements and exhibits structural compounding and superposition corresponding to different stages. The formation of surface tracks of the folds and faults has been affected by multidirectional extrusion stresses of the near SN, NE, and near EW. There are three stages of underground faults in the Longmaxi Formation, and the strikes are nearly EW, NE, and nearly SN. (2) Three distribution intervals for the homogenization temperature ranges of fracture fillings are 161–195°C, 121–143°C, and 74–105°C. The apatite thermal history simulation reveals that the Longmaxi Formation experienced three stages of tectonic movement after its formation. (3) There were clearly three stages in the structural development of the Longmaxi Formation in this area: the late Jurassic–Palaeocene (55 ± 5–38 ± 2 Ma), Eocene–early Miocene (38 ± 2–15.5 ± 3.5 Ma), and late Miocene-present (15.5 ± 3.5 Ma–present). Thus, a compound fracture system with superimposed structural deformations in different directions and at different stages formed in the study area. (4) A model for the stages and development of structural tracks in the Longmaxi Formation was established in conjunction with structural analysis and geomechanical theory. The results have guiding significance for the evaluation of shale gas preservation conditions and accumulation in the study area.

Dockerfile plays an important role in the Docker-based software development process, but many Dockerfile codes are infected with smells in practice. Understanding the occurrence of Dockerfile smells in open-source software can benefit the practice of Dockerfile and enhance project maintenance. In this paper, we perform an empirical study on a large dataset of 6,334 projects to help developers gain some insights into the occurrence of Dockerfile smells, including its coverage, distribution, co-occurrence, and correlation with project characteristics. Our results show that smells are very common in Dockerfile codes and there exists co-occurrence between different types of Dockerfile smells. Further, using linear regression analysis, when controlled for various variables, we statistically identify and quantify the relationships between Dockerfile smells occurrence and project characteristics. We also provide a rich resource of implications for software practitioners.

Summary: Flexible biosensors have received intensive attention for real-time, non-invasive monitoring of cancer biomarkers. Highly sensitive tyrosinase biosensors, which are important for melanoma screening, remained a hurdle. Herein, high-performance tyrosinase-sensing field-effect transistor-based biosensors (bio-FETs) have been successfully achieved by self-assembling nanostructured tetrapeptide tryptophan–valine–phenylalanine–tyrosine (WVFY) on n-type metal oxide transistors. In the presence of target tyrosinase, the phenolic hydroxyl groups in WVFY are rapidly converted to benzoquinone with the consumption of protons, which could be detected potentiometrically by bio-FETs. As a result, the WVFY-modified bio-FETs exhibited an ultra-low detection limit of 1.9 fM and an optimal detection range of 10 fM to 1 nM toward tyrosinase sensing. Furthermore, flexible devices fabricated on ∼2.9-μm-thick polyimide (PI) substrates illustrated robust mechanical flexibility, which could be attached to human skin conformally. These achievements hold promise for wearable melanoma screening and provide designing guidelines for detecting other important cancer biomarkers with bio-FETs.

Online learning platforms provide an opportunity to better understand students' weaknesses by tracking both their learning behavior and knowledge. This information can then be used to assist teachers in making instructional decisions and to further guide those who are at risk of failure. In this paper, we tracked student learning data from a C++ programming course over a whole semester of their freshman year via the Trustie platform. A total of 17,854 code submissions were collected. We then used CppCheck, SonarQube and Trustie to capture the quality characteristics and submission characteristics of the code, including lineOfCode, cyclomaticComplexity, codeSmell, syntacticError, averageScore, submission, and logicError, and analyzed the impact of code quality on the assignment work results. Several factors were discovered that we believe can help teachers to develop more effective teaching strategies.

Large-gap clones, a kind of clones that reuses code with many edits, are very common in software development practice and widespread in software systems. The detection of such clones is very crucial. However, due to a large number of edits, most of the current work fails to detect such clones effectively. This paper aims to find an effective approach for accurate detection of large-gap clones. We transform the code clone detection problem into a biological sequence alignment question and propose a novel approach that combines code fingerprint with sequence alignment. The sequence alignment is Smith-Waterman algorithm based, but shows significant improvements using dynamic parameter acquisition strategy. Furthermore, we design new rational criteria for clone identification. The proposed approach is automatically evaluated extensively by more than 10 million lines of code for general clones detection. We further conduct an empirical study on five large-scale Java projects to manually measure the approach for large-gap clones detection. The experimental results show that the proposed approach can effectively detect large-gap clones and exhibit good performance, and at the same time remains the competitiveness with existing advanced detection tools in detecting general clone detection.

Summary: Intracellular protein filaments are ubiquitous for cellular functions, but forming bona fide biomimetic intracellular filaments of small molecules in living cells remains elusive. Here, we report the in situ formation of self-limiting intracellular filaments of a small peptide via enzymatic morphological transition of a phosphorylated and trimethylated heterochiral tetrapeptide. Enzymatic dephosphorylation reduces repulsive intermolecular electrostatic interactions and converts the peptidic nanoparticles into filaments, which exhibit distinct types of cross-β structures with either C7 or C2 symmetries, with the hydrophilic C-terminal residues at the periphery of the helix. Macromolecular crowding promotes the peptide filaments to form bundles, which extend from the plasma membrane to nuclear membrane and hardly interact with endogenous components, including cytoskeletons. Stereochemistry and post-translational modification (PTM) of peptides are critical for generating the intracellular bundles. This work may offer a way to gain lost functions or to provide molecular insights for understanding normal and aberrant intracellular filaments.

In recent years, hundreds of novel small molecular drugs used for different treatments have been studied in the three phases of clinical trials around the world. However, less than 10% of them are eventually used due to diverse problems. Even some traditional drugs that have been approved by the Food and Drug Administration (FDA) have faced similar dilemmas. For instance, many drugs have poor water solubility, are easily hydrolyzed, or possess undesirable toxicity, while a variety of cancer cells develop drug resistance (DR) or multiple drug resistance (MDR) towards chemotherapeutic agents after long-term therapy. In order to improve the efficacy and efficiency of drugs, research has been directed forward towards the creation of assemblies of peptide–drug conjugates (PDCs) which have proven to possess wide potential for overcoming such complications based on their excellent biocompatibility, controllable biodegradability, site-selective targeting, and comparably low cytotoxicity. In this review, we focus on the recent developments and advances made in the creation of self-assembled nanostructures of PDCs for cancer therapy, on the chemical and physical properties of such drugs and peptides, and how they are arranged together to form diverse supramolecular nanostructures. Additionally, we cover certain mechanisms regarding how peptides or their derivatives enhance the efficiency and efficacy of those selected drugs and provide a brief discussion regarding the perspectives and remaining challenges in this intriguing field.

Recently, deep reinforcement learning (RL) algorithms have achieved significant progress in the multi-agent domain. However, training for increasingly complex tasks would be time-consuming and resource intensive. To alleviate this problem, efficient leveraging of historical experience is essential, which is under-explored in previous studies because most existing methods fail to achieve this goal in a continuously dynamic system owing to their complicated design. In this paper, we propose a method for knowledge reuse called “KnowRU”, which can be easily deployed in the majority of multi-agent reinforcement learning (MARL) algorithms without requiring complicated hand-coded design. We employ the knowledge distillation paradigm to transfer knowledge among agents to shorten the training phase for new tasks while improving the asymptotic performance of agents. To empirically demonstrate the robustness and effectiveness of KnowRU, we perform extensive experiments on state-of-the-art MARL algorithms in collaborative and competitive scenarios. The results show that KnowRU outperforms recently reported methods and not only successfully accelerates the training phase, but also improves the training performance, emphasizing the importance of the proposed knowledge reuse for MARL.

The advance of structural biology has revealed numerous noncovalent interactions between peptide sequences in protein structures, but such information is less explored for developing peptide materials. Here we report the formation of heterotypic peptide hydrogels by the two binding motifs revealed by the structures of an inflammasome. Specifically, conjugating a self-assembling motif to the positively or negatively charged peptide sequence from the ASCPYD filaments of inflammasome produces the solutions of the peptides. The addition of the peptides of the oppositely charged and complementary peptides to the corresponding peptide solution produces the heterotypic hydrogels. Rheology measurement shows that ratios of the complementary peptides affect the viscoelasticity of the resulted hydrogel. Circular dichroism indicates that the addition of the complementary peptides results in electrostatic interactions that modulate self-assembly. Transmission electron microscopy reveals that the ratio of the complementary peptides controls the morphology of the heterotypic peptide assemblies. This work illustrates a rational, biomimetic approach that uses the structural information from the protein data base (PDB) for developing heterotypic peptide materials via self-assembly.

In a decentralized multi-robot exploration problem, the robots have to cooperate effectively to map a strange environment as soon as possible without a centralized controller. In the past few decades, a set of “human-designed” cooperation strategies have been proposed to address this problem, such as the well-known frontier-based approach. However, many real-world settings, especially the ones that are constantly changing, are too complex for humans to design efficient and decentralized strategies. This paper presents a novel approach, the Attention-based Communication neural network (CommAttn), to “learn” the cooperation strategies automatically in the decentralized multi-robot exploration problem. The communication neural network enables the robots to learn the cooperation strategies with explicit communication. Moreover, the attention mechanism we introduced additionally can precisely calculate whether the communication is necessary for each pair of agents by considering the relevance of each received message, which enables the robots to communicate only with the necessary partners. The empirical results on a simulated multi-robot disaster exploration scenario demonstrate that our proposal outperforms the traditional “human-designed” methods, as well as other competing “learning-based” methods in the exploration task.

A biocompatible method of glutathione (GSH) catalyzed disulfide bond reduction was used to form Fmoc-short peptide-based supramolecular hydrogels. The hydrogels could form in both buffer solution and cell culture medium containing 10% of Fetal Bovine Serum (FBS) within minutes. The hydrogel was characterized by rheology, transmission electron microscopy, and fluorescence emission spectra. Their potential in three dimensional (3D) cell culture was evaluated and the results indicated that the gel with a low concentration of the peptide (0.1 wt%) was suitable for 3D cell culture of 3T3 cells. This study provides an alternative candidate of supramolecular hydrogel for 3D cell culture and cell delivery.