The protein sequence of PEP, accession number ACI03586.2, was obtained from the National Centre for Biotechnology Information (NCBI) database to perform analysis. Sequences from GenBank, RefSeq, TPA, SwissProt, PIR, PRF, and PDB are available in the comprehensive NCBI Protein database (http://www.ncbi.nlm.nih.gov/protein). Numerous analyses and investigations were made feasible by the NCBI Protein database's retrieval of the particular protein sequence, ACI03586.2. Annotated translations of coding regions from GenBank, RefSeq, and TPA, along with records from SwissProt, PIR, PRF, and PDB, are all available in the Protein database. The wealth of protein sequences gathered from these varied sources forms the essential foundation for comprehending the structure and function of living things.

The PEP protein was analyzed using the NCBI BLAST tool (http://blast.ncbi.nlm.nih.gov/Blast.cgi), more especially the Basic Local Alignment Search Tool (BLAST). The PEP protein sequence was compared to a non-redundant protein database as part of this analysis. This BLAST analysis aimed to find similar regions and possible matches with other proteins in the database. Additionally, we looked into the PEP protein's possible conserved domains during our analysis. The parts of the protein sequence known as conserved domains, which show a high degree of similarity between various proteins, can reveal important information about the functional properties and evolutionary links of the protein. The NCBI BLAST tool was an invaluable resource for these investigations as it provided us with the capability we needed.

In order to find putative homologous structures, we used the NCBI BLAST tool and, more specifically, performed a PSI-BLAST analysis using the query protein sequence. The protein data bank (PDB) served as the target database for this analysis. By using PSI-BLAST, we sought to find proteins in the PDB that show a significant degree of sequence similarity to our query protein. This method allows for the identification of putative homologous structures, which can offer insights into the three-dimensional architecture and possible functions of the query protein. By combining the strength of PSI-BLAST with the PDB's richness, our objective was to shed light on the putative homologous structures of the query protein, which can help us understand its potential biological.

Four protein sequences of PEP, which showed several features including the highest total score > 1000, E-value = 0, and identity > 80% retrieved from the previous step were aligned for homology evaluation. For analysis of the validity of the same sequences, the amino acid sequence of Prolyl-endopeptidase was used for alignment against template sequences based on an already conducted template search. All alignment generations were performed using the CLC 6.1 software. The BLOSUM substitution matrix was selected with values of 10, and 0.1, respectively for gap penalty gap extension penalty.

We used two online databases, TOPCONS and TMHMM, to identify the protein's hydrophobic transmembrane region. TOPCONS is a consensus prediction tool specifically made for membrane protein topology and signal peptide analysis. It may be accessed at https://topcons.cbr.su.se. When the protein's amino acid sequence is entered into TOPCON in FASTA format, it produces predictions about the membrane topology and finds putative signal peptides. We also used the TMHMM database, which is accessible at https://services.healthtech.dtu.dk/service.php, in addition to TOPCONS.TMHMM-2.0. Transmembrane Hidden Markov Model, or TMHMM for short, is a popular technique for identifying transmembrane helices in protein sequences. TMHMM uses a statistical model to detect putative transmembrane regions and determine how hydrophobic they are based on the protein sequence.

Our approach of choice for predicting protein secondary structures was the self-optimized prediction method (SOPM). We used SOPM (http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl page = npsa_sopma.html ) to anticipate secondary structural components including alpha helices, beta strands, and coils by analyzing the protein sequence. Using a combination of statistical models and algorithms, SOPM is a computational technique that generates precise predictions based on the amino acid sequence.

Three approaches were used for modeling, which included SWISS-MODEL (available at https://swissmodel.expasy.org), (PS)2 at http://ps.life.nctu.edu.tw/index.php, Phyre2 (accessible via the online address: http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgiid=index, and I-TASSER (Iterative Threading ASSEmbly Refinement) (available at https://zhanggroup.org/ I TASSER). The final 3D structure was built by the modeling package MODELLER.

To evaluate the models, we utilized QMEAN (Qualitative Model Energy Analysis), a widely used tool for assessing the quality of protein structures. QMEAN provides valuable insights into the major geometrical features of protein structures and helps to estimate their overall reliability and accuracy. The QMEAN analysis was performed using the online platform available at http://swissmodel.expasy.org/qmean/cgi/index.cgi, which offers a comprehensive set of evaluation parameters and scoring functions (21).

ModRefiner, an advanced tool accessible at http://zhanglab.ccmb.med.umich.edu/ModRefiner/, played a pivotal role in the refinement process of our models. This software specializes in constructing and refining protein structures starting from Cα traces using atomic-level energy minimization. By utilizing ModRefiner, we were able to enhance the accuracy and physical realism of our initial models. The refinement process involved iteratively adjusting the positions of both backbone and side-chain atoms, allowing for comprehensive flexibility during the simulations. This flexible conformational search was guided by a combination of physics-based and knowledge-based force fields, enabling the models to converge toward their native state (22).

To identify the best structural alignment and compare the predicted 3D structure of the protein with a template exhibiting favorable features, we employed the Dali web-based tool available at http://ekhidna.biocenter.helsinki.fi/dali_server/. Firstly, we selected the best-predicted 3D structure generated for the protein in our study. This structure was chosen based on its overall quality, reliability, and adherence to known structural principles. Subsequently, we utilized the Dali web-based tool to perform structural alignment between the selected 3D structure and the template with the most desirable characteristics. The template, identified by its PDB code as 3qlB, exhibited satisfactory features that made it a suitable reference for comparison.

In our study, we employed COFACTOR, a widely used tool available at http://zhanglab.ccmb.med.umich.edu/COFACTOR/, for the prediction of ligand binding sites. COFACTOR is a comprehensive method that integrates structure, sequence, and protein-protein interaction information to annotate the biological roles of proteins. Using COFACTOR, we aimed to identify and characterize the specific regions within the protein structure that are involved in binding to ligands. Ligand binding sites play crucial roles in protein function, as they are responsible for interactions with small molecules, ions, or other proteins, often leading to important biological activities.

we utilized two important tools for the analysis of protein structures. The first tool, InterProSurf, available at http://curie.utmb.edu/prosurf.html, was employed to identify interface residues in protein complexes. InterProSurf is a powerful server that specializes in predicting interacting sites on protein surfaces. By analyzing protein complex structures deposited in the Protein Data Bank (PDB), InterProSurf enables the identification of residues involved in protein-protein interactions. This information is valuable for understanding the functional and structural aspects of protein complexes. The second tool we utilized is GHECOM (Grid-based HECOMi finder), accessible at https://pdbj.org/ghecom/. GHECOM is a program designed to identify multi-scale pockets on protein surfaces using mathematical morphology. It employs a grid-based approach to detect these pockets, which can be potential binding sites for ligands or other molecules. By analyzing the protein surface at different scales, GHECOM allows the identification of various types of pockets, including small, medium, and large ones. This information provides insights into the structural characteristics and potential functional roles of these pockets.

The consurf program (http://consurf.tau.ac.il/) was used for the identification of functional residues of the nanobody structure. The software parameters were set as PSI-BLAST for five iterations against the Uniprot database with an E-value of 0.01 and maximum likelihood (ML) to calculate the amino acid conservation score.

For achievement of information like the protein packing, residue volume, contacts, accessibility, and topological genus calculations, VLDPws based on the Laguerre diagram (https://www.dsimb.inserm.fr/dsimb_tools/vldp/index.php) is a routine tool, which was utilized in this study.

The BLAST search conducted using the query sequence yielded several significant hits, indicating sequence similarities with various organisms. Among the hits, notable matches included [Eurygaster integriceps], [Halyomorpha halys], [Apolygus lucorum], [Cimex lectularius], [Zootermopsis nevadensis], and several others. This finding suggests that the query sequence shares homology with these organisms, indicating potential evolutionary relationships or functional similarities. The identified hits represent species from diverse taxonomic groups, suggesting that the sequence may have conserved regions that are important for its biological function. Further analysis of the query sequence revealed the presence of putative conserved domains associated with the Prolyl oligopeptidase PreP. Conserved domains are specific regions within a protein sequence that are highly conserved across different species and are often associated with particular functions or structural features. In this case, the identified conserved domains are characteristic of the Prolyl oligopeptidase PreP protein.

Table 1 represents the first 10 hits having the highest BLAST scores on the query sequence against the PDB database. The first hit (Accession: 6JCI_A, the Crystal structure of PEP from Haliotis discus hannai with SUAM-14746 [Haliotis discus hannai], Max score: 734, Query coverage: 99%, Max ident: 50.35%) showed the best score and so was selected as a template.

When analyzing the sequence homology between PEP and the selected template, a schematic illustration in Fig. 1 was used to visually represent the comparison. The results of this comparison revealed that PEP and the template share a sequence identity of 50.35%. Sequence homology refers to the degree of similarity or identity between two protein sequences. In this case, PEP and the template were aligned and compared against each other, examining the matching residues and their positions. The sequence identity of 50.35% indicates that approximately half of the amino acid residues in PEP align with equivalent positions in the template sequence.

To gain insights into the structural organization of PEP, a topology model was constructed. The topology model provides information about the arrangement of different regions within the protein, including the presence or absence of transmembrane segments. In this case, the topology model for PEP indicated the absence of any transmembrane regions within the protein (Fig. 2). The absence of transmembrane regions suggests that PEP is likely a soluble protein, rather than being embedded within cell membranes. Soluble proteins are typically found in the cytoplasm, nucleus, or other cellular compartments where they perform various enzymatic or regulatory functions.

the evaluation of the protein secondary structure revealed the presence of multiple structural elements, including alpha helices, extended strands, beta-turns, and random coils. The proportions of these secondary structure components were determined, with alpha helix, extended strand, beta-turn, and random coil comprising approximately 24.75%, 26.44%, 8.02%, and 40.79% of the structure, respectively. The graphical representation in Fig. 3 visually depicts the distribution of these secondary structure elements, providing a comprehensive overview of the protein's structural composition.

Swiss model and ps2v2 recruited for homology modeling created two models. Phyre2 predicted one 3D model and I-TASSER suggested 5 models. I-TASSER simulations created a high number ensemble of structural conformations termed decoys. To screen the estimated models, I-TASSER employs the SPICKER program for clustering all the decoys based on the similarity of pair-wise structures and suggests several models up to five numbers corresponding to the five largest structure clusters. The confidence of models was quantitatively measured by calculation of the C-score according to the significance of threading template alignments and the convergence parameters of the structure assembly simulations. The normal range of the C-score is [-5, 2], where a higher value indicates a model with higher confidence and vice-versa. Moreover, the TM-score and RMSD (root-mean-square deviation) were calculated based on the C-score, and the protein length is determined by the association between these qualities. The model with the higher C-score (C-score = 1.42, Estimated TM-score = 0.91 ± 0.06, Estimated RMSD = 5.0 ± 3.3Å) was selected for further analysis.

The QMEAN server was employed to estimate the 3D models of the protein. Among the predicted models generated by different methods, including phyre2, Ps2v2, and I-Tasser, the Swiss Model 3D structure obtained the highest QMEAN score, indicating its superior quality. The graphical representation in Fig. 4 and the data in Table 2 serve to visually and quantitatively demonstrate the comparison of QMEAN scores for the different predicted models. The selection of the Swiss Model 3D structure based on its higher QMEAN score suggests that it is the most reliable and accurate model for further analysis.

The 3D structure of the protein initially existed as the primary model, which served as an initial approximation. Through a refinement process, the final model was obtained, representing an improved version with enhanced structural quality. The graphical representation in Fig. 5 visually displays the primary and final models, allowing for a direct comparison of the structural improvements achieved through refinement.

No significant difference was found between the structures built for the template and the 2D and 3D structures of the query protein. The comparison of their tertiary structures revealed matches in alignment, indicating a close similarity in the overall folding patterns and spatial arrangements of the secondary structure elements. The graphical representation in Fig. 6 visually demonstrates the alignment of the tertiary structures, further supporting the observed similarities between the template and the query protein.

The COFACTOR algorithm was utilized to predict the ligand binding sites within the protein. The analysis identified several conserved residues, including positions 173, 472, 475, 553, 554, 599, 644, 645, and 681, that are likely involved in ligand binding. The graphical representation in Fig. 7 visually illustrates the predicted ligand binding sites and highlights the positions of the conserved residues within these sites. This information provides valuable insights into the protein's functional characteristics and potential for ligand interactions.

The Interprosurf tool was utilized to find interface residues within the protein complex structure. Residues 566, 567, 568, 569, and 570 were predicted as interface residues and are depicted in Fig. 8. Additionally, the GHECOM server identified five pockets on the PEP surfaces of the protein complex, and these pockets are shown in Fig. 9. The information obtained from these analyses contributes to a better understanding of the protein complex's interactions, potential binding sites, and functional characteristics, facilitating further investigations in the field of molecular biology and drug discovery.

Consurf analysis was conducted to identify functional residues within the nanobody. Both the sequence and structure of the nanobody were analyzed to assess the degree of conservation and identify functionally important positions. The results of this analysis are presented in Fig. 10, offering a visual representation of the functional residues within the nanobody. These findings contribute to a better understanding of the nanobody's functional characteristics and can guide further experimental investigations and applications in various fields, such as antibody engineering and therapeutics.

The VLDP server was used to generate a contact map depicting the interactions between residues within the protein. The contact map in Fig. 11 provides a visual representation of these interactions. Additionally, the quantification of the contacts, represented by a blue scale legend, offers a numerical measure of the common area between the atoms of each residue. Furthermore, the VLDP server computed additional information about the contacts, including exposed surface area, total surface area, neighboring residues, sequence positions, and amino acid types. These details enhance the understanding of the contact patterns and contribute to the characterization of the protein's structure and functionality.

The Laguerre diagram, utilizing finely tuned parameters, enables the determination of accurate volume values for each residue within the protein. Table 3 presents the volume measurements for individual residues, providing insights into their spatial extent. The interactive nature of the volume data allows for dynamic exploration and analysis. These volume measurements contribute to a comprehensive understanding of the protein's structure and can be utilized to investigate its functional implications and molecular interactions.

The LOCUS column provides information about the nature of residue contacts, distinguishing between buried, solvent-exposed, and boundary residues. This information aids in understanding the accessibility and potential functional roles of residues within the protein. The PIA analysis, depicted in Fig. 12 and presented in Table 4, provides a detailed characterization of the residue contacts, including interaction types and strengths. Integrating these findings enhances the understanding of the protein's structure, interactions, and potential functional mechanisms.