Predictive modeling for breast cancer classification in the context of Bangladeshi patients by use of machine learning approach with explainable AI.

Breast cancer has rapidly increased in prevalence in recent years, making it one of the leading causes of mortality worldwide. Among all cancers, it is by far the most common. Diagnosing this illness manually requires significant time and expertise. Since detecting breast cancer is a time-consuming process, preventing its further spread can be aided by creating machine-based forecasts. Machine learning and Explainable AI are crucial in classification as they not only provide accurate predictions but also offer insights into how the model arrives at its decisions, aiding in the understanding and trustworthiness of the classification results. In this study, we evaluate and compare the classification accuracy, precision, recall, and F1 scores of five different machine learning methods using a primary dataset (500 patients from Dhaka Medical College Hospital). Five different supervised machine learning techniques, including decision tree, random forest, logistic regression, naive bayes, and XGBoost, have been used to achieve optimal results on our dataset. Additionally, this study applied SHAP analysis to the XGBoost model to interpret the model's predictions and understand the impact of each feature on the model's output. We compared the accuracy with which several algorithms classified the data, as well as contrasted with other literature in this field. After final evaluation, this study found that XGBoost achieved the best model accuracy, which is 97%.

Exploring the substrate spectrum of phylogenetically distinct bacterial polyesterases.

The rapid escalation of plastic waste accumulation presents a significant threat of the modern world, demanding an immediate solution. Over the last years, utilization of the enzymatic machinery of various microorganisms has emerged as an environmentally friendly asset in tackling this pressing global challenge. Thus, various hydrolases have been demonstrated to effectively degrade polyesters. Plastic waste streams often consist of a variety of different polyesters, as impurities, mainly due to wrong disposal practices, rendering recycling process challenging. The elucidation of the selective degradation of polyesters by hydrolases could offer a proper solution to this problem, enhancing the recyclability performance. Towards this, our study focused on the investigation of four bacterial polyesterases, including DaPUase, IsPETase, PfPHOase, and Se1JFR, a novel PETase-like lipase. The enzymes, which were biochemically characterized and structurally analyzed, demonstrated degradation ability of synthetic plastics. While a consistent pattern of polyesters' degradation was observed across all enzymes, Se1JFR stood out in the degradation of PBS, PLA, and polyether PU. Additionally, it exhibited comparable results to IsPETase, a benchmark mesophilic PETase, in the degradation of PCL and semi-crystalline PET. Our results point out the wide substrate spectrum of bacterial hydrolases and underscore the significant potential of PETase-like enzymes in polyesters degradation.

Cue-driven microbial cooperation and communication: evolving quorum sensing with honest signaling.

BACKGROUND: Quorum sensing (QS) is the ability of microorganisms to assess local clonal density by measuring the extracellular concentration of signal molecules that they produce and excrete. QS is also the only known way of bacterial communication that supports the coordination of within-clone cooperative actions requiring a certain threshold density of cooperating cells. Cooperation aided by QS communication is sensitive to cheating in two different ways: laggards may benefit from not investing in cooperation but enjoying the benefit provided by their cooperating neighbors, whereas Liars explicitly promise cooperation but fail to do so, thereby convincing potential cooperating neighbors to help them, for almost free. Given this double vulnerability to cheats, it is not trivial why QS-supported cooperation is so widespread among prokaryotes. RESULTS: We investigated the evolutionary dynamics of QS in populations of cooperators for whom the QS signal is an inevitable side effect of producing the public good itself (cue-based QS). Using spatially explicit agent-based lattice simulations of QS-aided threshold cooperation (whereby cooperation is effective only above a critical cumulative level of contributions) and three different (analytical and numerical) approximations of the lattice model, we explored the dynamics of QS-aided threshold cooperation under a feasible range of parameter values. We demonstrate three major advantages of cue-driven cooperation. First, laggards cannot wipe out cooperation under a wide range of reasonable environmental conditions, in spite of an unconstrained possibility to mutate to cheating; in fact, cooperators may even exclude laggards at high cooperation thresholds. Second, lying almost never pays off, if the signal is an inevitable byproduct (i.e., the cue) of cooperation; even very cheap fake signals are selected against. And thirdly, QS is most useful if local cooperator densities are the least predictable, i.e., if their lattice-wise mean is close to the cooperation threshold with a substantial variance. CONCLUSIONS: Comparing the results of the four different modeling approaches indicates that cue-driven threshold cooperation may be a viable evolutionary strategy for microbes that cannot keep track of past behavior of their potential cooperating partners, in spatially viscous and in well-mixed environments alike. Our model can be seen as a version of the famous greenbeard effect, where greenbeards coexist with defectors in a evolutionarily stable polymorphism. Such polymorphism is maintained by the condition-dependent trade-offs of signal production which are characteristic of cue-based QS.

Metabolism-triggered sensor array aided by machine learning for rapid identification of pathogens.

Chemical-nose strategy has achieved certain success in the discrimination and identification of pathogens. However, this strategy usually relies on non-specific interactions, which are prone to be significantly disturbed by the change of environment thus limiting its practical usefulness. Herein, we present a novel chemical-nose sensing approach leveraging the difference in the dynamic metabolic variation during peptidoglycan metabolism among different species for rapid pathogen discrimination. Pathogens were first tethered with clickable handles through metabolic labeling at two different acidities (pH = 5 and 7) for 20 and 60 min, respectively, followed by click reaction with fluorescence up-conversion nanoparticles to generate a four-dimensional signal output. This discriminative multi-dimensional signal allowed eight types of model bacteria to be successfully classified within the training set into strains, genera, and Gram phenotypes. As the difference in signals of the four sensing channels reflects the difference in the amount/activity of enzymes involved in metabolic labeling, this strategy has good anti-interference capability, which enables precise pathogen identification within 2 h with 100% accuracy in spiked urinary samples and allows classification of unknown species out of the training set into the right phenotype. The robustness of this approach holds significant promise for its widespread application in pathogen identification and surveillance.

Metabolic Modulation of Kynurenine Based on Kynureninase-Loaded Nanoparticle Depot Overcomes Tumor Immune Evasion in Cancer Immunotherapy.

Evading recognition of immune cells is a well-known strategy of tumors used for their survival. One of the immune evasion mechanisms is the synthesis of kynurenine (KYN), a metabolite of tryptophan, which suppresses the effector T cells. Therefore, lowering the KYN concentration can be an efficient antitumor therapy by restoring the activity of immune cells. Recently, kynureninase (KYNase), which is an enzyme transforming KYN into anthranilate, was demonstrated to show the potential to decrease KYN concentration and inhibit tumor growth. However, due to the limited bioavailability and instability of proteins in vivo, it has been challenging to maintain the KYNase concentration sufficiently high in the tumor microenvironment (TME). Here, we developed a nanoparticle system loaded with KYNase, which formed a Biodegradable and Implantable Nanoparticle Depot named 'BIND' following subcutaneous injection. The BIND sustainably supplied KYNase around the TME while located around the tumor, until it eventually degraded and disappeared. As a result, the BIND system enhanced the proliferation and cytokine production of effector T cells in the TME, followed by tumor growth inhibition and increased mean survival. Finally, we showed that the BIND carrying KYNase significantly synergized with PD-1 blockade in three mouse models of colon cancer, breast cancer, and melanoma.

Genomic characterization and related functional genes of γ- poly glutamic acid producing Bacillus subtilis.

γ- poly glutamic acid (γ-PGA), a high molecular weight polymer, is synthesized by microorganisms and secreted into the extracellular space. Due to its excellent performance, γ-PGA has been widely used in various fields, including food, biomedical and environmental fields. In this study, we screened natto samples for two strains of Bacillus subtilis N3378-2at and N3378-3At that produce γ-PGA. We then identified the γ-PGA synthetase gene cluster (PgsB, PgsC, PgsA, YwtC and PgdS), glutamate racemase RacE, phage-derived γ-PGA hydrolase (PghB and PghC) and exo-γ-glutamyl peptidase (GGT) from the genome of these strains. Based on these γ-PGA-related protein sequences from isolated Bacillus subtilis and 181 B. subtilis obtained from GenBank, we carried out genotyping analysis and classified them into types 1-5. Since we found B. amyloliquefaciens LL3 can produce γ-PGA, we obtained the B. velezensis and B. amyloliquefaciens strains from GenBank and classified them into types 6 and 7 based on LL3. Finally, we constructed evolutionary trees for these protein sequences. This study analyzed the distribution of γ-PGA-related protein sequences in the genomes of B. subtilis, B. velezensis and B. amyloliquefaciens strains, then the evolutionary diversity of these protein sequences was analyzed, which provided novel information for the development and utilization of γ-PGA-producing strains.

Efficient Conversion of Stevioside to Rebaudioside M in Saccharomyces cerevisiae by a Engineering Hydrolase System and Prolonging the Growth Cycle.

Rebaudioside (Reb) M is an important sweetener with high sweetness, but its low content in Stevia rebaudiana and low catalytic capacity of the glycosyltransferases in heterologous microorganisms limit its production. In order to improve the catalytic efficiency of the conversion of stevioside to Reb M by Saccharomyces cerevisiae, several key issues must be resolved including knocking out endogenous hydrolases, enhancing glycosylation, and extending the enzyme catalytic process. Herein, endogenous glycosyl hydrolase SCW2 was knocked out in S. cerevisiae. The glycosylation process was enhanced by screening glycosyltransferases, and UGT91D2 from S. rebaudiana was identified as the optimum glycosyltransferase. The UDP-glucose supply was enhanced by overexpressing UGP1, and co-expressing UGT91D2 and UGT76G1 achieved efficient conversion of stevioside to Reb M. In order to extend the catalytic process, the silencing information regulator 2 (SIR2) which can prolong the growth cycle of S. cerevisiae was introduced. Finally, combining these modifications produced 12.5 g/L Reb M and the yield reached 77.9% in a 5 L bioreactor with 10.0 g/L stevioside, the highest titer from steviol glycosides to Reb M reported to date. The engineered strain could facilitate the industrial production of Reb M, and the strategies provide references for the production of steviol glycosides.

Identification and Biological Evaluation of a Novel Small-Molecule Inhibitor of Ricin Toxin.

The plant-derived toxin ricin is classified as a type 2 ribosome-inactivating protein (RIP) and currently lacks effective clinical antidotes. The toxicity of ricin is mainly due to its ricin toxin A chain (RTA), which has become an important target for drug development. Previous studies have identified two essential binding pockets in the active site of RTA, but most existing inhibitors only target one of these pockets. In this study, we used computer-aided virtual screening to identify a compound called RSMI-29, which potentially interacts with both active pockets of RTA. We found that RSMI-29 can directly bind to RTA and effectively attenuate protein synthesis inhibition and rRNA depurination induced by RTA or ricin, thereby inhibiting their cytotoxic effects on cells in vitro. Moreover, RSMI-29 significantly reduced ricin-mediated damage to the liver, spleen, intestine, and lungs in mice, demonstrating its detoxification effect against ricin in vivo. RSMI-29 also exhibited excellent drug-like properties, featuring a typical structural moiety of known sulfonamides and barbiturates. These findings suggest that RSMI-29 is a novel small-molecule inhibitor that specifically targets ricin toxin A chain, providing a potential therapeutic option for ricin intoxication.

Developing an Improved Cycle Architecture for AI-Based Generation of New Structures Aimed at Drug Discovery.

Drug discovery involves a crucial step of optimizing molecules with the desired structural groups. In the domain of computer-aided drug discovery, deep learning has emerged as a prominent technique in molecular modeling. Deep generative models, based on deep learning, play a crucial role in generating novel molecules when optimizing molecules. However, many existing molecular generative models have limitations as they solely process input information in a forward way. To overcome this limitation, we propose an improved generative model called BD-CycleGAN, which incorporates BiLSTM (bidirectional long short-term memory) and Mol-CycleGAN (molecular cycle generative adversarial network) to preserve the information of molecular input. To evaluate the proposed model, we assess its performance by analyzing the structural distribution and evaluation matrices of generated molecules in the process of structural transformation. The results demonstrate that the BD-CycleGAN model achieves a higher success rate and exhibits increased diversity in molecular generation. Furthermore, we demonstrate its application in molecular docking, where it successfully increases the docking score for the generated molecules. The proposed BD-CycleGAN architecture harnesses the power of deep learning to facilitate the generation of molecules with desired structural features, thus offering promising advancements in the field of drug discovery processes.

An empirical assessment of the influence of digital transformation on sports corporate sustainability.

The trend of digital transformation fosters enterprise change, helps cultivate enterprises' own competitive advantages and is crucial to the advancement of sports enterprises' sustainable development in the framework of the emerging digital economy as a national strategy. However, there have been few empirical studies on the microlevel of digital transformation and its impact on the sustainability of sports organizations. Therefore, the sustainable growth dynamic model is used to construct indicators of corporate sustainability by referencing 48 sports corporations listed on Shanghai and Shenzhen A-shares markets and the New Third Board in China from 2012 to 2021. The intrinsic relationship between digital transformation and the sustainable development of sports enterprises and the underlying mechanism of action are explored by constructing a panel fixed effects model, a chain mediating effects model, and a panel threshold model. The most important contribution is as follows: To provide a useful reference for analyzing enterprise digital transformation, a more complete indicator indicating the extent of corporate digital transformation is built. The micro viewpoint broadens our awareness of sustainable development in sports organizations and deepens our understanding of the interaction model between sustainable development and enterprise digital transformation. This study provides methodical evidence and insights for an accurate understanding of digital transformation for sustainable enterprise development, looking into the "black box" of the mechanism between digital transformation and sustainable business development. The results show that digital transformation significantly aids sports enterprises in their pursuit of long-term sustainability. Heterogeneity tests demonstrate the pivotal role of digital transformation in advancing the sustained growth of sports firms and high-tech sports enterprises situated in the eastern region of China. Regarding transmission mechanisms, the chain mediating effect of enterprises' digital transformation on improved technological innovation and TFP, which in turn promote long-term business growth, has yet to be validated. Further examination exposes that within the context of the correlation between digital transformation and the sustainability of corporations, there is a single threshold effect based on financing restrictions and operational costs and a double threshold effect based on operational efficiency.

Semantic segmentation of urban environments: Leveraging U-Net deep learning model for cityscape image analysis.

Semantic segmentation of cityscapes via deep learning is an essential and game-changing research topic that offers a more nuanced comprehension of urban landscapes. Deep learning techniques tackle urban complexity and diversity, which unlocks a broad range of applications. These include urban planning, transportation management, autonomous driving, and smart city efforts. Through rich context and insights, semantic segmentation helps decision-makers and stakeholders make educated decisions for sustainable and effective urban development. This study investigates an in-depth exploration of cityscape image segmentation using the U-Net deep learning model. The proposed U-Net architecture comprises an encoder and decoder structure. The encoder uses convolutional layers and down sampling to extract hierarchical information from input images. Each down sample step reduces spatial dimensions, and increases feature depth, aiding context acquisition. Batch normalization and dropout layers stabilize models and prevent overfitting during encoding. The decoder reconstructs higher-resolution feature maps using "UpSampling2D" layers. Through extensive experimentation and evaluation of the Cityscapes dataset, this study demonstrates the effectiveness of the U-Net model in achieving state-of-the-art results in image segmentation. The results clearly shown that, the proposed model has high accuracy, mean IOU and mean DICE compared to existing models.

Identification of Novel Antimicrobial Compounds Targeting Mycobacterium tuberculosis S-Adenosyl-L-Homocysteine Hydrolase Using Dual Hierarchical In Silico Structure-Based Drug Screening.

The emergence of multidrug-resistant and extensively drug-resistant Mycobacterium tuberculosis (M. tuberculosis) has become a major medical problem. S-adenosyl-L-homocysteine hydrolase (MtSAHH) was selected as the target protein for the identification of novel anti-TB drugs. Dual hierarchical in silico Structure-Based Drug Screening was performed using a 3D compound structure library (with over 150 thousand synthetic chemicals) to identify compounds that bind to MtSAHH's active site. In vitro experiments were conducted to verify whether the nine compounds selected as new drug candidates exhibited growth-inhibitory effects against mycobacteria. Eight of the nine compounds that were predicted by dual hierarchical screening showed growth-inhibitory effects against Mycobacterium smegmatis (M. smegmatis), a model organism for M. tuberculosis. Compound 7 showed the strongest antibacterial activity, with an IC50 value of 30.2 µM. Compound 7 did not inhibit the growth of Gram-negative bacteria or exert toxic effects on human cells. Molecular dynamics simulations of 40 ns using the MtSAHH-Compound 7 complex structure suggested that Compound 7 interacts stably with the MtSAHH active site. These in silico and in vitro results suggested that Compound 7 is a promising lead compound for the development of new anti-TB drugs.

Molecular Insights into the Enhanced Activity and/or Thermostability of PET Hydrolase by D186 Mutations.

PETase exhibits a high degradation activity for polyethylene terephthalate (PET) plastic under moderate temperatures. However, the effect of non-active site residues in the second shell of PETase on the catalytic performance remains unclear. Herein, we proposed a crystal structure- and sequence-based strategy to identify the key non-active site residue. D186 in the second shell of PETase was found to be capable of modulating the enzyme activity and stability. The most active PETaseD186N improved both the activity and thermostability with an increase in Tm by 8.89 °C. The PET degradation product concentrations were 1.86 and 3.69 times higher than those obtained with PETaseWT at 30 and 40 °C, respectively. The most stable PETaseD186V showed an increase in Tm of 12.91 °C over PETaseWT. Molecular dynamics (MD) simulations revealed that the D186 mutations could elevate the substrate binding free energy and change substrate binding mode, and/or rigidify the flexible Loop 10, and lock Loop 10 and Helix 6 by hydrogen bonding, leading to the enhanced activity and/or thermostability of PETase variants. This work unraveled the contribution of the key second-shell residue in PETase in influencing the enzyme activity and stability, which would benefit in the rational design of efficient and thermostable PETase.

HINT2 protects against pressure overload-induced cardiac remodelling through mitochondrial pathways.

Histidine triad nucleotide-binding protein 2 (HINT2) is an enzyme found in mitochondria that functions as a nucleotide hydrolase and transferase. Prior studies have demonstrated that HINT2 plays a crucial role in ischemic heart disease, but its importance in cardiac remodelling remains unknown. Therefore, the current study intends to determine the role of HINT2 in cardiac remodelling. HINT2 expression levels were found to be lower in failing hearts and hypertrophy cardiomyocytes. The mice that overexpressed HINT2 exhibited reduced myocyte hypertrophy and cardiac dysfunction in response to stress. In contrast, the deficiency of HINT2 in the heart of mice resulted in a worsening hypertrophic phenotype. Further analysis indicated that upregulated genes were predominantly associated with the oxidative phosphorylation and mitochondrial complex I pathways in HINT2-overexpressed mice after aortic banding (AB) treatment. This suggests that HINT2 increases the expression of NADH dehydrogenase (ubiquinone) flavoprotein (NDUF) genes. In cellular studies, rotenone was used to disrupt mitochondrial complex I, and the protective effect of HINT2 overexpression was nullified. Lastly, we predicted that thyroid hormone receptor beta might regulate HINT2 transcriptional activity. To conclusion, the current study showcased that HINT2 alleviates pressure overload-induced cardiac remodelling by influencing the activity and assembly of mitochondrial complex I. Thus, targeting HINT2 could be a novel therapeutic strategy for reducing cardiac remodelling.

[Structure motif guided mining of MHET hydrolase and development of a two-enzyme cascade for plastics depolymerization at mild temperature].

The utilization of polyethylene terephthalate (PET) has caused significant and prolonged ecological repercussions. Enzymatic degradation is an environmentally friendly approach to addressing PET contamination. Hydrolysis of mono(2-hydroxyethyl) terephthalate (MHET), a competitively inhibited intermediate in PET degradation, is catalyzed by MHET degrading enzymes. Herein, we employed bioinformatic methods that combined with sequence and structural information to discover an MHET hydrolase, BurkMHETase. Enzymatic characterization showed that the enzyme was relatively stable at pH 7.5-10.0 and 30-45 ℃. The kinetic parameters kcat and Km on MHET were (24.2±0.5)/s and (1.8±0.2) µmol/L, respectively, which were similar to that of the well-known IsMHETase with higher substrate affinity. BurkMHETase coupled with PET degradation enzymes improved the degradation of PET films. Structural analysis and mutation experiments indicated that BurkMHETase may have evolved specific structural features to hydrolyze MHET. For MHET degrading enzymes, aromatic amino acids at position 495 and the synergistic interactions between active sites or distal amino acids appear to be required for MHET hydrolytic activity. Therefore, BurkMHETase may have substantial potential in a dual-enzyme PET degradation system while the bioinformatic methods can be used to broaden the scope of applicable MHETase enzymes.

[Whole-cell catalytic production of pseudouridine by recombinant Escherichia coli].

Pseudouridine is the most abundant modified nucleoside found in non-coding RNA and is widely used in biological and pharmaceutical fields. However, current methods for pseudouridine production suffer from drawbacks such as complex procedures, low efficiency and high costs. This study presents a novel enzymatic cascade reaction route in Escherichia coli, enabling the whole-cell catalytic synthesis of pseudouridine from uridine. Initially, a metabolic pathway was established through plasmid-mediated overexpression of endogenous pseudouridine-5-phosphase glycosidase, ribokinase, and ribonucleoside hydrolase, resulting in the accumulation of pseudouridine. Subsequently, highly active endogenous ribonucleoside hydrolase was screened to enhance uridine hydrolysis and provide more precursors for pseudouridine synthesis. Furthermore, modifications were made to the substrates and products transport pathways to increase the pseudouridine yield while avoiding the accumulation of by-product uridine. The resulting recombinant strain Ψ-7 catalyzed the conversion of 30 g/L uridine into 27.24 g/L pseudouridine in 24 h, achieving a conversion rate of 90.8% and a production efficiency of 1.135 g/(L·h). These values represent the highest reported yield and production efficiency achieved by enzymatic catalysis methods to date.

Enhancing Point-of-Care Diagnosis of African Swine Fever Virus (ASFV) DNA with the CRISPR-Cas12a-Assisted Triplex Amplified Assay.

Accurate, ultrasensitive, and point-of-care (POC) diagnosis of the African swine fever virus (ASFV) remains imperative to prevent its spread and limit the losses incurred. Herein, we propose a CRISPR-Cas12a-assisted triplex amplified colorimetric assay for ASFV DNA detection with ultrahigh sensitivity and specificity. The specific recognition of recombinase aided amplification (RAA)-amplified ASFV DNA could activate the Cas12a/crRNA/ASFV DNA complex, leading to the digestion of the linker DNA (bio-L1) on magnetic beads (MBs), thereby preventing its binding of gold nanoparticles (AuNPs) network. After magnetic separation, the release of AuNPs network comprising a substantial quantity of AuNPs could lead to a discernible alteration in color and significantly amplify the plasmonic signal, which could be read by spectrophotometers or smartphones. By combining the RAA, CRISPR/Cas12a-assisted cleavage, and AuNPs network-mediated colorimetric amplification together, the assay could detect as low as 0.1 copies/µL ASFV DNA within 1 h. The assay showed an accuracy of 100% for the detection of ASFV DNA in 16 swine tissue fluid samples, demonstrating its potential for on-site diagnosis of ASFV.

Cybersecurity on a budget: Evaluating security and performance of open-source SIEM solutions for SMEs.

The proliferation of cyber threats necessitates robust security measures to safeguard critical assets and data in today's evolving digital landscape. Small and Medium Enterprises (SMEs), which are the backbone of the global economy are particularly vulnerable to these threats due to inadequate protection for critical and sensitive information, budgetary constraints, and lack of cybersecurity expertise and personnel. Security Information and Event Management (SIEM) systems have emerged as pivotal tools for monitoring, detecting, and responding to security incidents. While proprietary SIEM solutions have historically dominated the market, open-source SIEM systems have gained prominence for their accessibility and cost-effectiveness for SMEs. This article presents a comprehensive study focusing on the evaluation of open-source SIEM systems. The research investigates the capabilities of these open-source solutions in addressing modern security challenges and compliance with regulatory requirements. Performance aspects are explored through empirical testing in simulated enterprise-grade SME network environments to assess resource utilization, and real-time data processing capabilities. By providing a rigorous assessment of the security and performance features of open-source SIEM systems, this research offers valuable insights to cybersecurity practitioners, organizations seeking cost-effective security solutions, and the broader academic community. The findings shed light on the strengths and limitations of these systems, aiding decision-makers in selecting the most suitable SIEM solution for their specific requirements while enhancing the cybersecurity posture of SMEs.

Characterization, Semirational Design for pH Robustness, and the Application in Bioactive Peptide Production of a X-Prolyl Dipeptidyl Aminopeptidase from Lactococcus lactis MY-3.

PepXLcMY-3, an X-prolyl dipeptidyl aminopeptidase derived from Lactobacillus lactis MY-3, was screened and recombinantly expressed in Escherichia coli. The enzyme could exhibit about 40% activity within the pH range of 6.0-10. To further improve the pH robustness, site E396 located in the active pocket was discovered through alanine scanning. The mutant E396I displayed both developed activity and kcat/Km. The optimal pH of E396I shifted from 6.0 to 10 compared to WT, with the relative activity within the pH range of 6.0-10 significantly increased. The site K648 was then proposed by semirational design. The activity of mutant E396I/K648D reached 4.03 U/mg. The optimal pH was restored to 6.0, and the pH stability was further improved. E396I/K648D could totally hydrolyze ß-casomorphin 7 within 30 min. The hydrolysate showed 64.5% inhibition on angiotensin I converting enzyme, which was more efficient than those produced by E396I and WT, 23.2 and 44.7%, respectively.

Machine learning-aided inverse design for biogas upgrading through biological CO2 conversion.

The biogas upgrading process through bioconversion of CO2 to CH4 by hydrogenotrophic methanogens is an attractive strategy for energy decarbonation. Many studies have optimized operational parameters to improve key performance indicators such as CH4% and H2 utilization efficiency. However, inconsistent laboratory conditions make it challenging to compare results. Existing models for analyzing operating conditions can only assess the impact of individual conditions and lack the ability to simultaneously optimize multiple conditions. To address this, two XGBoost models were built with R2 of 0.779 and 0.903 with data collected from literatures and were embedded into multi-objective partitive swarm optimization algorithm to optimal operating conditions. Predictions were compared with experimental validations under optimized conditions, revealing an 8.50% and 2.95% relative error in CH4% and H2 conversion rate, respectively. This approach streamlines biogas upgrading processes, offering a data-driven solution to enhance efficiency and consistency in the pursuit of sustainable methane production.