Identification Of Proteins Research Articles

TurboID mapping reveals the exportome of secreted intrinsically disordered proteins in the transforming parasite Theileria annulata.

Theileria annulata is a tick-transmitted apicomplexan parasite that gained the unique ability among parasitic eukaryotes to transform its host cell, inducing a fatal cancer-like disease in cattle. Understanding the mechanistic interplay between the host cell and malignant Theileria species that drives this transformation requires the identification of responsible parasite effector proteins. In this study, we used TurboID-based proximity labeling, which unbiasedly identified secreted parasite proteins within host cell compartments. By fusing TurboID to nuclear export or localization signals, we biotinylated proteins in the vicinity of the ligase enzyme in the nucleus or cytoplasm of infected macrophages, followed by mass spectrometry analysis. Our approach revealed with high confidence nine nuclear and four cytosolic candidate parasite proteins within the host cell compartments, eight of which had no orthologs in non-transforming T. orientalis. Strikingly, all eight of these proteins are predicted to be highly intrinsically disordered proteins. We discovered a novel tandem arrayed protein family, nuclear intrinsically disordered proteins (NIDP) 1-4, featuring diverse functions predicted by conserved protein domains. Particularly, NIDP2 exhibited a biphasic host cell-cycle-dependent localization, interacting with the EB1/CD2AP/CLASP1 parasite membrane complex at the schizont surface and the tumor suppressor stromal antigen 2 (STAG2), a cohesion complex subunit, in the host nucleus. In addition to STAG2, numerous NIDP2-associated host nuclear proteins implicated in various cancers were identified, shedding light on the potential role of the T. annulata exported protein family NIDP in host cell transformation and cancer-related pathways.IMPORTANCETurboID proximity labeling was used to identify secreted proteins of Theileria annulata, an apicomplexan parasite responsible for a fatal, proliferative disorder in cattle that represents a significant socio-economic burden in North Africa, central Asia, and India. Our investigation has provided important insights into the unique host-parasite interaction, revealing secreted parasite proteins characterized by intrinsically disordered protein structures. Remarkably, these proteins are conspicuously absent in non-transforming Theileria species, strongly suggesting their central role in the transformative processes within host cells. Our study identified a novel tandem arrayed protein family, with nuclear intrinsically disordered protein 2 emerging as a central player interacting with established tumor genes. Significantly, this work represents the first unbiased screening for exported proteins in Theileria and contributes essential insights into the molecular intricacies behind the malignant transformation of immune cells.

Proteomic Identification of Plasma Proteins Associated with Future Dementia Development.

Proteomic Identification of Plasma Proteins Associated with Future Dementia Development.

Understanding bacterial pathogen diversity: A proteogenomic analysis and use of an array of genome assemblies to identify novel virulence factors of the honey bee bacterial pathogen Paenibacillus larvae.

Mass spectrometry proteomics data are typically evaluated against publicly available annotated sequences, but the proteogenomics approach is a useful alternative. A single genome is commonly utilized in custom proteomic and proteogenomic data analysis. We pose the question of whether utilizing numerous different genome assemblies in a search database would be beneficial. We reanalyzed raw data from the exoprotein fraction of four reference Enterobacterial Repetitive Intergenic Consensus (ERIC) I-IV genotypes of the honey bee bacterial pathogen Paenibacillus larvae and evaluated them against three reference databases (from NCBI-protein, RefSeq, and UniProt) together with an array of protein sequences generated by six-frame direct translation of 15 genome assemblies from GenBank. The wide search yielded 453 protein hits/groups, which UpSet analysis categorized into 50 groups based on the success of protein identification by the 18 database components. Nine hits that were not identified by a unique peptide were not considered for marker selection, which discarded the only protein that was not identified by the reference databases. We propose that the variability in successful identifications between genome assemblies is useful for marker mining. The results suggest that various strains of P. larvae can exhibit specific traits that set them apart from the established genotypes ERIC I-V.

Iron-sulfur protein odyssey: exploring their cluster functional versatility and challenging identification.

Iron-sulfur (Fe-S) clusters are an essential and ubiquitous class of protein-bound prosthetic centers that are involved in a broad range of biological processes (e.g. respiration, photosynthesis, DNA replication and repair and gene regulation) performing a wide range of functions including electron transfer, enzyme catalysis, and sensing. In a general manner, Fe-S clusters can gain or lose electrons through redox reactions, and are highly sensitive to oxidation, notably by small molecules such as oxygen and nitric oxide. The [2Fe-2S] and [4Fe-4S] clusters, the most common Fe-S cofactors, are typically coordinated by four amino acid side chains from the protein, usually cysteine thiolates, but other residues (e.g. histidine, aspartic acid) can be found. While diversity in cluster coordination ensures the functional variety of the Fe-S clusters, the lack of conserved motifs makes new Fe-S protein identification challenging especially when the Fe-S cluster is also shared between two proteins as observed in several dimeric transcriptional regulators and in the mitoribosome. Thanks to the recent development of in cellulo, in vitro and in silico approaches, new Fe-S proteins are still regularly identified, highlighting the functional diversity of this class of proteins. In this review, we will present three main functions of the Fe-S clusters and explain the difficulties encountered to identify Fe-S proteins and methods that have been employed to overcome these issues.

Identification and Quantification of Human Relaxin Proteins by Immunoaffinity-Mass Spectrometry.

The human relaxins belong to the Insulin/IGF/Relaxin superfamily of peptide hormones, and their physiological function is primarily associated with reproduction. In this study, we focused on a prostate tissue-specific relaxin RLN1 (REL1_HUMAN protein) and a broader tissue specificity RLN2 (REL2_HUMAN protein). Due to their structural similarity, REL1 and REL2 proteins were collectively named a 'human relaxin protein' in previous studies and were exclusively measured by immunoassays. We hypothesized that the highly selective and sensitive immunoaffinity-selected reaction monitoring (IA-SRM) assays would reveal the identity and abundance of the endogenous REL1 and REL2 in biological samples and facilitate the evaluation of these proteins for diagnostic applications. High levels of RLN1 and RLN2 transcripts were found in prostate and breast cancer cell lines by RT-PCR. However, no endogenous prorelaxin-1 or mature REL1 were detected by IA-SRM in cell lines, seminal plasma, or blood serum. The IA-SRM assay of REL2 demonstrated its undetectable levels (<9.4 pg/mL) in healthy control female and male sera and relatively high levels of REL2 in maternal sera across different gestational weeks (median 331 pg/mL; N = 120). IA-SRM assays uncovered potential cross-reactivity and nonspecific binding for relaxin immunoassays. The developed IA-SRM assays will facilitate the investigation of the physiological and pathological roles of REL1 and REL2 proteins.

Ion Moblity Mass Spectrometry in Multi-omics Studies

Mass spectrometry (MS) as an analytical technique enables the identification and quantitative determination of proteins, metabolites, or lipids in a studied sample. However, this method has limitations regarding the number of molecules that can be identified at a given time. To increase the number of identifications, the application of ion mobility spectrometry (IMS) can be employed. This technique allows the separation of ions based on their mobility while traversing the analyser in a gradient of an electromagnetic field and opposing gas pressure. The separation is performed in conjunction with MS analysis, adding another dimension to the analysis, resulting in a significant improvement in the number of identified compounds and a reduction in noise. Alternatively, while maintaining the same number of identifications, analysis can be performed in a shorter time period. It is crucial to pay special attention to the type of IMS analyser used, as its specific implementation dictates further stages of analysis and ion detection capabilities.

Super-resolution proximity labeling with enhanced direct identification of biotinylation sites

Promiscuous labeling enzymes, such as APEX2 or TurboID, are commonly used in in situ biotinylation studies of subcellular proteomes or protein–protein interactions. Although the conventional approach of enriching biotinylated proteins is widely implemented, in-depth identification of specific biotinylation sites remains challenging, and current approaches are technically demanding with low yields. A novel method to systematically identify specific biotinylation sites for LC-MS analysis followed by proximity labeling showed excellent performance compared with that of related approaches in terms of identification depth with high enrichment power. The systematic identification of biotinylation sites enabled a simpler and more efficient experimental design to identify subcellular localized proteins within membranous organelles. Applying this method to the processing body (PB), a non-membranous organelle, successfully allowed unbiased identification of PB core proteins, including novel candidates. We anticipate that our newly developed method will replace the conventional method for identifying biotinylated proteins labeled by promiscuous labeling enzymes.

SLC34A2 promotes cell proliferation by activating STX17-mediated autophagy in esophageal squamous cell carcinoma.

Solute carrier family 34 member 2 (SLC34A2) has been implicated in the development of various malignancies. However, the clinical significance and underlying molecular mechanisms of SLC34A2 in esophageal squamous cell carcinoma (ESCC) remain elusive. Western blotting, quantitative real-time PCR and immunohistochemistry were utilized to evaluate the expression levels of SLC34A2 mRNA/protein in ESCC cell lines or tissues. Kaplan-Meier curves were employed for survival analysis. CCK-8, colony formation, EdU and xenograft tumor model assays were conducted to determine the impact of SLC34A2 on ESCC cell proliferation. Cell cycle was examined using flow cytometry. RNA-sequencing and enrichment analysis were carried out to explore the potential signaling pathways. The autophagic flux was evaluated by western blotting, mRFP-GFP-LC3 reporter system and transmission electron microscopy. Immunoprecipitation and mass spectrometry were utilized for identification of potential SLC34A2-interacting proteins. Cycloheximide (CHX) chase and ubiquitination assays were conducted to test the protein stability. The expression of SLC34A2 was significantly upregulated in ESCC and correlated with unfavorable clinicopathologic characteristics particularly the Ki-67 labeling index and poor prognosis of ESCC patients. Overexpression of SLC34A2 promoted ESCC cell proliferation, while silencing SLC34A2 had the opposite effect. Moreover, SLC34A2 induced autophagy to promote ESCC cell proliferation, whereas inhibition of autophagy suppressed the proliferation of ESCC cells. Further studies showed that SLC34A2 interacted with an autophagy-related protein STX17 to promote autophagy and proliferation of ESCC cells by inhibiting the ubiquitination and degradation of STX17. These findings indicate that SLC34A2 may serve as a prognostic biomarker for ESCC.

Milk Fat Globule Proteins Are Relevant Bovine Milk Allergens in Patients with α‐Gal Syndrome

Alpha-gal syndrome (AGS) is a mammalian meat allergy associated with tick bites and specific IgE to the oligosaccharide galactose-α-1,3-galactose (α-gal). Recent studies have shown that 10-20% of AGS patients also react to the dairy proteins. Considering the already described role of the meat lipid fraction in AGS manifestations, the aim of this work has been to investigate whether the milk fat globule proteins (MFGPs) could be involved in AGS. The MFGPs are extracted and their recognition by the IgE of AGS patients is proved through immunoblotting experiments. The identification of the immunoreactive proteins by LC-HRMS analysis allows to demonstrate for the first time that butyrophillin, lactadherin, and xanthine oxidase (XO) are α-gal glycosylated. The role of xanthine oxidase seems to be prevalent since it is highly recognized byboth the anti-α-gal antibody and AGS patient sera. The results obtained in this study provide novel insights in the characterization of α-Gal carrying glycoproteins in bovine milk, supporting the possibility that milk, especially in its whole form, may give reactions in AGS patients. Although additional factors are probably associated with the clinical manifestations, the avoidance of milk and milk products should be considered in individuals with AGS showing symptoms related to milk consumption.

Characterization of effective, simple, and low-cost precipitation methods for depleting abundant plasma proteins to enhance the depth and breadth of plasma proteomics.

Plasma is an abundant source of proteins and potential biomarkers to aid in the detection, diagnosis, and prognosis of human diseases. These proteins are often present at low levels in the blood and difficult to identify and measure due to the large dynamic range of proteins. The goal of this work was to characterize and compare various protein precipitation methods related to how they affect the depth and breadth of plasma proteomic studies. Abundant protein precipitation with perchloric acid (PerCA) can increase protein identifications and depth of plasma proteomic studies. Three acid- and four solvent-based precipitation methods were evaluated. All methods tested provided excellent plasma proteomic coverage (>600 identified protein groups) and detected protein in the low pg/mL range. Functional enrichment analysis revealed subtle differences within and larger changes between the precipitant groups. Methanol-based precipitation outperformed the other methods based on identifications and reproducibility. The methods' performance was verified using eight lung cancer patient samples, where >700 protein groups were measured and proteins with an estimated plasma concentration of ∼10pg/mL were detected. Various protein precipitation agents are amenable to extending the depth and breadth of plasma proteomes. These data can guide investigators to implement inexpensive, high-throughput methods for their plasma proteomic workflows.

Recent Progress in the Application of Tau Protein Biosensors for Diagnosis of Neurodegenerative Diseases.

The microtubule-associated Tau protein is found in the central nervous system (CNS) in six major isoforms. Neurodegenerative diseases have been linked to post-translational changes of Tau, most notably phosphorylation. Tau protein's molecular diversity is highly helpful in the identification of neurodegenerative illnesses. Nonetheless, one major obstacle to the early detection of brain illness is the nanoscale identification of tau proteins. The standard methods for identifying tau protein include western blotting, polymerase chain reaction (PCR), and real-time PCR. Enzyme-linked immunosorbent assay (ELISA) is another approach used. The limited sensitivity and specificity of these detections, together with the need for sophisticated equipment, are some of their drawbacks. The development of innovative and complex methods for tau protein screening is necessary to address the aforementioned issues. Biosensors are a cutting-edge instrument that may help identify various neurodegenerative biomarkers as early as feasible. This paper provides an overview of the most recent developments in the detection of neurodegenerative diseases employing biosensors built on nanotechnology and methods for imaging, electrochemical, and optical detection of the Tau protein. Furthermore, we outline the present difficulties and suggest a possible course for biosensor-based detection and intervention in the future.

Identification and Characterization of a Bacterial Periplasmic Solute Binding Protein That Binds l-Amino Acid Amides.

Periplasmic solute-binding proteins (SBPs) are key ligand recognition components of bacterial ATP-binding cassette (ABC) transporters that allow bacteria to import nutrients and metabolic precursors from the environment. Periplasmic SBPs comprise a large and diverse family of proteins, of which only a small number have been empirically characterized. In this work, we identify a set of 610 unique uncharacterized proteins within the SBP_bac_5 family that are found in conserved operons comprising genes encoding (i) ABC transport systems and (ii) putative amidases from the FmdA_AmdA family. From these uncharacterized SBP_bac_5 proteins, we characterize a representative periplasmic SBP from Mesorhizobium sp. A09 (MeAmi_SBP) and show that MeAmi_SBP binds l-amino acid amides but not the corresponding l-amino acids. An X-ray crystal structure of MeAmi_SBP bound to l-serinamide highlights the residues that impart distinct specificity for l-amino acid amides and reveals a structural Ca2+ binding site within one of the lobes of the protein. We show that the residues involved in ligand and Ca2+ binding are conserved among the 610 SBPs from experimentally uncharacterized FmdA_AmdA amidase-associated ABC transporter systems, suggesting these homologous systems are also likely to be involved in the sensing, uptake, and metabolism of l-amino acid amides across many Gram-negative nitrogen-fixing soil bacteria. We propose that MeAmi_SBP is involved in the uptake of such solutes to supplement pathways such as the citric acid cycle and the glutamine synthetase-glutamate synthase pathway. This work expands our currently limited understanding of microbial interactions with l-amino acid amides and bacterial nitrogen utilization.

Regulatory sequence-based discovery of anti-defense genes in archaeal viruses

In silico identification of viral anti-CRISPR proteins (Acrs) has relied largely on the guilt-by-association method using known Acrs or anti-CRISPR associated proteins (Acas) as the bait. However, the low number and limited spread of the characterized archaeal Acrs and Aca hinders our ability to identify Acrs using guilt-by-association. Here, based on the observation that the few characterized archaeal Acrs and Aca are transcribed immediately post viral infection, we hypothesize that these genes, and many other unidentified anti-defense genes (ADG), are under the control of conserved regulatory sequences including a strong promoter, which can be used to predict anti-defense genes in archaeal viruses. Using this consensus sequence based method, we identify 354 potential ADGs in 57 archaeal viruses and 6 metagenome-assembled genomes. Experimental validation identified a CRISPR subtype I-A inhibitor and the first virally encoded inhibitor of an archaeal toxin-antitoxin based immune system. We also identify regulatory proteins potentially akin to Acas that can facilitate further identification of ADGs combined with the guilt-by-association approach. These results demonstrate the potential of regulatory sequence analysis for extensive identification of ADGs in viruses of archaea and bacteria.

A methodological exploration of distinguishing hair quality based on hair proteomics

Hair is an advantageous biological sample due to its recordable, collectable, and storable nature. Hair's primary components are keratin and keratin-associated proteins. Owing to its abundance of cystine, keratin possesses impressive mechanical strength and chemical stability, formed by creating disulfide bonds as crosslinks within the protein peptide chain. Furthermore, keratin is cross-linked with keratin-associated proteins to create a complex network structure that provides the hair with strength and rigidity. Protein extraction serves as the foundation for hair analysis research. Bleaching hair causes damage to the structure between keratin and keratin-associated proteins, resulting in texture issues and hair breakage. This article outlines various physical treatment methods and lysate analysis that enhance the efficiency of hair protein extraction. The PLEE method achieves a three-fold increase in hair protein extraction efficiency when using a lysis solution containing SDS and combining high temperatures with intense shaking, compared to previous methods found in literature. We utilized the PLEE method to extract hair from both normal and damaged groups. Normal samples identified 156–157 proteins, including 51 keratin and keratin-associated proteins. The damaged group consisted of 155–158 identified proteins, of which 48–50 were keratin and keratin-associated proteins. Bleaching did not cause any notable difference in the protein identification of hair. However, it did reduce coverage of keratin and keratin-associated proteins significantly. Our hair protein extraction method provides extensive coverage of the hair proteome. Our findings indicate that bleaching damage results in subpar hair quality due to reduced coverage of protein primary sequences in keratin and keratin-associated proteins.

Increasing protein identifications in bottom-up proteomics of T. castaneum − Exploiting synergies of protein biochemistry and bioinformatics

Depending on the respective research question, LC-MS/MS based bottom-up proteomics poses challenges from the initial biological sample all the way to data evaluation. The focus of this study was to investigate the influence of sample preparation techniques and data analysis parameters on protein identification in Tribolium castaneum by applying free software proteomics platform Max Quant.Multidimensional protein extraction strategies in combination with electrophoretic or chromatographic off-line protein pre-fractionation were applied to enhance the spectrum of isolated proteins from T. castaneum and reduce the effect of co-elution and ion suppression effects during nano-LC-MS/MS measurements of peptides. For comprehensive data analysis, MaxQuant was used for protein identification and R for data evaluation. A wide range of parameters were evaluated to gain reproducible, reliable, and significant protein identifications.A simple phosphate buffer, pH 8, containing protease and phosphatase inhibitor cocktail and application of gentle extraction conditions were used as a first extraction step for T.castaneum proteins. Furthermore, a two-dimensional extraction procedure in combination with electrophoretic pre-fractionation of extracted proteins and subsequent in-gel digest resulted in almost 100 % increase of identified proteins when compared to chromatographic fractionation as well as one-pot-analysis. The additionally identified proteins could be assigned to new molecular functions or cell compartments, emphasizing the positive effect of extended sample preparation in bottom-up proteomics. Besides the number of peptides during post-processing, MaxQuant’s Match between Runs exhibited a crucial effect on the number of identified proteins. A maximum relative standard deviation of 2 % must be considered for the data analysis.Our work with Tribolium castaneum larvae demonstrates that sometimes – depending on matrix and research question − more complex and time-consuming sample preparation can be advantageous for isolation and identification of additional proteins in bottom-up proteomics.

Identification of nitrated proteins in human thrombi from ischemic cerebrovascular events

Identification of nitrated proteins in human thrombi from ischemic cerebrovascular events

Fermentation Kinetics, Potability, and Sensory Compounds of Joubidwi: An Ethnic Alcoholic Beverage

Background: Fermentation is a metabolic pathway of the breakdown of carbohydrate molecules like glucose, fructose to ethanol and various volatile and non-volatile congeners. The production of alcoholic beverages from foodstuffs like rice, cereals, fruits, maize etc., is an age-old traditional protocol. In this study, we adapted the traditional protocol of preparation of the ethnic beverage Joubidwi, with a fermentation time of 6 days, i.e., aliquots from the fermented mash (cooked rice) were collected on a daily basis from day 1 to day 6. Objectives: This study aimed to get insight into fermentation kinetics, ethanol content, carbohydrate content, antioxidant activities, bitterness, protein content, pH, colour, and turbidity of the ethnic alcoholic beverage Joubidwi and to identify the sensory compounds and polyphenol molecules. Method: UV-Visible spectroscopic methods, GC-FID and GC-MS analytics methods were used for the estimation and identification of carbohydrates, proteins, ethanol contents, and sensory compounds, respectively. Reverse phase HPLC was adapted for the identification of polyphenol molecules in Joubidwi. Results: A pseudo-first-order fermentation kinetics was observed from day 1 to day 6 fermentation of Joubidwi. Apart from ethanol content (60.48 ± 1.36 mg/mL after 72 h fermentation to 78.80 ± 1.48 mg/mL after 144 h fermentation), the beverage had nutritional values and good antioxidant properties with 22.38 to 35.11% Radical Scavenging Effect (RSE) from 1 to 6th day of fermented Joubidwi. The beverage also contained 12 sensory compounds. Gallic acid was identified as polyphenol in Joubidwi samples. Conclusion: Joubidwi is an alcoholic beverage prepared by Bodos using rice and certain herbs through a traditional fermentation process. The fermentation process follows pseudo-first-order kinetics. The beverage is potable.

Improved Data Acquisition Settings on Q Exactive HF-X and Fusion Lumos Tribrid Orbitrap-Based Mass Spectrometers for Proteomic Analysis of Limited Samples.

Deep proteomic profiling of complex biological and medical samples available at low nanogram and subnanogram levels is still challenging. Thorough optimization of settings, parameters, and conditions in nanoflow liquid chromatography-tandem mass spectrometry (MS)-based proteomic profiling is crucial for generating informative data using amount-limited samples. This study demonstrates that by adjusting selected instrument parameters, e.g., ion injection time, automated gain control, and minimally altering the conditions for resuspending or storing the sample in solvents of different compositions, up to 15-fold more thorough proteomic profiling can be achieved compared to conventionally used settings. More specifically, the analysis of 1 ng of the HeLa protein digest standard by Q Exactive HF-X Hybrid Quadrupole-Orbitrap and Orbitrap Fusion Lumos Tribrid mass spectrometers yielded an increase from 1758 to 5477 (3-fold) and 281 to 4276 (15-fold) peptides, respectively, demonstrating that higher protein identification results can be obtained using the optimized methods. While the instruments applied in this study do not belong to the latest generation of mass spectrometers, they are broadly used worldwide, which makes the guidelines for improving performance desirable to a wide range of proteomics practitioners.

Exploring Lung Cancer Protein Network: Understanding Structure and Function through Metric Space Modeling

Lung cancer remains a significant health threat with high mortality rates. Using graph theory, we modeled the protein-protein interaction (PPI) network in lung cancer to explore its complex structure. This approach allows for the analysis of network properties and the identification of key proteins driving biological processes. Our analysis revealed RPS27A as a central proteinwithin the network, associated with diverse functions related to ribosome biogenesis, translation, cell growth, apoptosis, and cancer progression. This suggests that RPS27A may have multiple functions in cancer development and progression, including in MAPK signaling pathways. Importantly, our study uniquely identifies RPS27A as a central hub in lung cancer PPI networks, shedding light on its pivotal role in disease pathogenesis. Additionally, we identified a central network zone enriched with proteins involved in key signaling pathways, presenting novel insights into potential therapeutic targets for lung cancer treatment. Pathway enrichment analysis further highlighted functional specialization across network zones, providing a comprehensive understanding of the intricate interplay between biological pathways in lung cancer progression. This study underscores the multifaceted roles of central proteins like RPS27A within lung cancer’s PPI network and the network’s potential for pinpointing therapeutic targets, presenting a novel perspective on the intricate network of molecular interactions driving lung cancer pathogenesis.

Model fusion for predicting unconventional proteins secreted by exosomes using deep learning.

Unconventional secretory proteins (USPs) are vital for cell-to-cell communication and are necessary for proper physiological processes. Unlike classical proteins that follow the conventional secretory pathway via the Golgi apparatus, these proteins are released using unconventional pathways. The primary modes of secretion for USPs are exosomes and ectosomes, which originate from the endoplasmic reticulum. Accurate and rapid identification of exosome-mediated secretory proteins is crucial for gaining valuable insights into the regulation of non-classical protein secretion and intercellular communication, as well as for the advancement of novel therapeutic approaches. Although computational methods based on amino acid sequence prediction exist for predicting unconventional proteins secreted by exosomes (UPSEs), they suffer from significant limitations in terms of algorithmic accuracy. In this study, we propose a novel approach to predict UPSEs by combining multiple deep learning models that incorporate both protein sequences and evolutionary information. Our approach utilizes a convolutional neural network (CNN) to extract protein sequence information, while various densely connected neural networks (DNNs) are employed to capture evolutionary conservation patterns.By combining six distinct deep learning models, we have created a superior framework that surpasses previous approaches, achieving an ACC score of 77.46% and an MCC score of 0.5406 on an independent test dataset.