PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Prediction Algorithms”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 73 records · Page 4Linked to original sources
PubMed1991

Prediction of protein backbone conformation based on seven structure assignments. Influence of local interactions.

A method is developed to compute backbone tertiary folds from the amino acid sequence. In this method, the number of degrees of freedom is drastically reduced by neglecting side-chain flexibility, and by describing backbone conformations as combinations of only seven structural states. These are characterized by single values of the dihedral angles phi, psi and omega, representing allowed conformations of the isolated dipeptide. We show that this restrictive model is none the less capable of describing native backbones to within acceptable deviations. Using our backbone description, potentials of mean force are derived from a database of known protein structures, based on statistical influences of single residues and residue pairs on the conformational states in their vicinity along the chain. This yields the force-field component due to local interactions, which is then used to predict lowest-energy conformations from any given amino acid sequence. The prediction algorithm does not require searching conformational space and is therefore extremely fast. Another important asset of our method is that it is able to compute not only the minimum energy conformation, but any number of lowest energy structures, whose relative preferences can be determined from the corresponding computed energy values. The performance of our procedure is tested on short peptides that are likely to be stabilized by local interactions. These include several helical structures and a hexapeptide with a beta-bend conformation, corresponding to peptides shown to have relatively well-defined conformations in aqueous solution, and to protein segments believed to adopt their native conformation early during folding. In addition, several flexible peptides are analysed. Except for the problems encountered in predicting observed disulphide bridges in two of the flexible peptides, and in a somewhat larger fragment comprising residues 30 to 51 of bovine trypsin inhibitor, prediction results compare very favourably with experimental data. Potential applications of our procedure to protein modelling and its extension to protein folding are discussed.

Amino Acid Sequence↗
PubMed2025

Unraveling the complex genetic landscape of OTOF-related hearing loss: a deep dive into cryptic variants and haplotype phasing.

BACKGROUND: Pathogenic variants in OTOF are a major cause of auditory synaptopathy. However, challenges remain in interpreting OTOF variants, including difficulties in confirming haplotype phasing using traditional short-read sequencing (SRS) due to the large gene size, the potential incomplete penetrance of certain variants, and difficulties in assessing variants at non-canonical splice sites. This study aims to revisit the genetic landscape of OTOF variants in a Taiwanese non-syndromic auditory neuropathy spectrum disorder (ANSD) cohort using a combination of sequencing technologies, predictive tools, and experimental validations. METHODS: We performed SRS to analyze OTOF variants in 65 unrelated Taiwanese patients diagnosed with non-syndromic ANSD, complemented by long-read sequencing (LRS) for haplotype phasing. A prediction-to-validation pipeline was implemented to assess the pathogenicity of cryptic variants using SpliceAI software and minigene assays. RESULTS: Biallelic pathogenic OTOF variants were identified in 33 patients (50.8%), while monoallelic variants were found in five patients. Three novel variants, c.3864G > A (p.Ala1288 =), c.4501G > A (p.Ala1501Thr), and c.5813 + 2T > C, were detected. The pathogenicity of two non-canonical mis-splicing variants, c.3894 + 5G > C and c.3864G > A (p.Ala1288 =), was confirmed by minigene assays. LRS-based haplotype phasing revealed that the common missense variant c.5098G > C (p.Glu1700Gln) and the novel variant c.5975A > G (p.Lys1992Arg) are in cis and form a founder pathogenic allele in the Taiwanese population. CONCLUSIONS: Our study highlights the genetic heterogeneity of DFNB9 and emphasizes the importance of population-specific variant interpretation. The integration of advanced sequencing technologies, predictive algorithms, and functional validation assays will improve the accuracy of molecular diagnosis and inform personalized treatment strategies for individuals with DFNB9.

Humans↗
PubMed1990

Conformational analysis of the immunodominant epitopes of the circumsporozoite protein of Plasmodium falciparum and knowlesi.

All of the coat proteins of the sporozoite and merozoite stages of Plasmodium, determined to date, contain tandem repeats and most of these contain at least one proline residue. These tandemly repeated segments of the circumsporozite (CS) proteins of P. falciparum and P. knowlesi have been shown to constitute an immunodominant epitope. Antibodies to these peptide segments have been shown to be protective and cause the shedding of the CS protein, known as the CSP reaction. In this study, four synthetic peptides were prepared by solid-phase peptide synthesis. The first peptide corresponds to the tetrapeptide tandem repeat in the CS protein of P. falciparum, repeated eight times, (NANP)8. The second peptide is an analogue of the first in which glycine is substituted for proline, (NANG)8. The third peptide corresponds to the tandem repeat of P. knowlesi, PK(1-24), which is repeated twice (QAQGDGANAGQP)2. The fourth peptide is a tetrapeptide repeat, corresponding to the C-terminal tetrapeptide of PK(1-24) and is repeated eight times, (AGQP)8. It is shown by CD measurements that the presence of proline in these repeats induces an increase in beta-sheet (beta-turn) content in the (NANP)8 peptide relative to the repeat of (NANG)8 and PK(1-24) peptide in aqueous media. The (AGQP)8 peptide has the highest beta-sheet (beta-turn) content of all peptides studied. The Chou-Fasman predictive algorithm indicates a high beta-turn content in the synthetic peptides. It is concluded that this increase in defined structure correlates well with and hence may contribute to the increased antigenicity in these repeats.

Amino Acids↗
PubMed1992

Isolation and cloning of Omp alpha, a coiled-coil protein spanning the periplasmic space of the ancestral eubacterium Thermotoga maritima.

We have discovered a new oligomeric protein component associated with the outer membrane of the ancestral eubacterium Thermotoga maritima. In electron micrographs, the protein, Omp alpha, appears as a rod-shaped spacer that spans the periplasm, connecting the outer membrane to the inner cell body. Purification, biochemical characterization and sequencing of Omp alpha suggest that it is a homodimer composed of two subunits of 380 amino acids with a calculated M(r) of 43,000 and a pI of 4.54. The sequence of the omp alpha gene indicates a tripartite organization of the protein with a globular NH2-terminal domain of 64 residues followed by a putative coiled-coil segment of 300 residues and a COOH-terminal, membrane-spanning segment. The predicted length of the coiled-coil segment (45 nm) correlates closely with the spacing between the inner and outer membranes. Despite sequence similarity to a large number of coiled-coil proteins and high scores in a coiled-coil prediction algorithm, the sequence of the central rod-shaped domain of Omp alpha does not have the typical 3.5 periodicity of coiled-coil proteins but rather has a periodicity of 3.58 residues. Such a periodicity was also found in the central domain of staphylococcal M protein and beta-giardin and might be indicative of a subclass of fibrous proteins with packing interactions that are distinct from the ones seen in other two-stranded coiled-coils.

Amino Acid Sequence↗
PubMed1991

Peak assignment in automatic data analysis.

Linear prediction algorithms are able to identify peaks in an NMR spectrum, but are not able to assign these peaks to components anticipated in the spectrum. We have developed an artificial-intelligence protocol which uses the output parameter list from an LPSVD algorithm, and automatically assigns the peaks on the basis of an anticipated list of components. To overcome the influence of experimental conditions on the absolute values of frequency, integrated area, and linewidth, the assignment routine performs an internal scaling of the data by comparing all possible pairs of peaks in the spectrum. Completely automated analysis of large numbers of in vivo FIDs is now possible.

Adenosine Triphosphate↗
PubMed1990

Conformational flexibility of a scorpion toxin active on mammals and insects: a circular dichroism study.

Three scorpion toxins have been analyzed by circular dichroism in water and in 2,2,2-trifluoroethanol (TFE) solutions. These toxins were chosen because they are representative of three kinds of pharmacological activities: (1) toxin AaH IT2, an antiinsect toxin purified from the venom of Androctonus australis Hector, which is able to bind to insect nervous system preparation, (2) toxin Css II, from the venom of Centruroides suffusus suffusus, which is a beta-type antimammal toxin capable of binding to mammal nervous system preparation, and (3) the toxin Ts VII from the venom of Tityus serrulatus, which is able to bind to both types of nervous systems. In order to minimize bias, CD data were analyzed by a predictive algorithm to assess secondary structure content. Among the three molecules, Ts VII presented the most unordered secondary structure in water, but it gained in ordered forms when solubilized in TFE. These results indicated that the Ts VII backbone is the most flexible, which might result in a more pronounced tendency for this toxin molecule to undergo conformational changes. This is consistent with the fact that it competes with both antiinsect and beta-type antimammal toxins for the binding to the sodium channel.

Amino Acid Sequence↗
PubMed1992

A structural assessment of the apo[a] protein of human lipoprotein[a].

Apolipoprotein[a], the highly glycosylated, hydrophilic apoprotein of lipoprotein[a] (Lp[a]), is generally considered to be a multimeric homologue of plasminogen, and to exhibit atherogenic/thrombogenic properties. The cDNA-inferred amino acid sequence of apo[a] indicates that apo[a], like plasminogen and some zymogens, is composed of a kringle domain and a serine protease domain. To gain insight into possible positive functions of Lp[a], we have examined the apo[a] primary structure by comparing its sequence with those of other proteins involved in coagulation and fibrinolysis, and its secondary structure by using a combination of structure prediction algorithms. The kringle domain encompasses 11 distinct types of repeating units, 9 of which contain 114 residues. These units, called kringles, are similar but not identical to each other or to PGK4. Each apo[a] kringle type was compared with kringles which have been shown to bind lysine and fibrin, and with bovine prothrombin kringle 1. Apo[a] kringles are linked by serine/threonine- and proline-rich stretches similar to regions in immunoglobulins, adhesion molecules, glycoprotein Ib-alpha subunit, and kininogen. In comparing the protease domains of apo[a] and plasmin, apo[a] contains a region between positions 4470 and 4492 where 8 substitutions, 9 deletions, and 1 insertion are apparent. Our analysis suggests that apo[a] kringle-type 10 has a high probability of binding to lysine in the same way as PGK4. In the only human apo[a] polymorph sequenced to date, position 4308 is occupied by serine, whereas the homologous position in plasmin is occupied by arginine and is an important site for proteolytic cleavage and activation. An alternative site for the proteolytic activation of human apo[a] is proposed.

Amino Acid Sequence↗
PubMed1989

Helix formation in reduced, S-carboxymethylated human choriogonadotropin beta subunit and tryptic peptides.

The beta subunit of human choriogonadotropin (hCG beta) and its asialoderivative were digested with trypsin and then reduced and S-carboxymethylated. A series of peptides were purified which corresponded to residues 1-43, 44-95, 96-114, and 123-145 of the 145 amino acid residue glycoprotein. The two N-linked oligosaccharides were present on the amino terminal peptide, and three of the four O-linked oligosaccharides were present on the carboxy terminal peptide. Circular dichroic spectra between 190-240 nm were obtained on reduced, S-carboxymethylated (RCM) hCG beta and the above peptides, both in aqueous solution and in the helicogenic solvent 80% (vol/vol) trifluoroethanol (TFE). In aqueous solution there was evidence of only limited helicity in the peptides and RCM-hCG beta; however, in the presence of TFE, peptides 1-43 and 44-95 exhibited significant helicity, as did the full-length linear chain. The helicity developed in TFE by RCM-hCG beta appears much greater than that which occurs in the native, disulfide-intact form, thus suggesting that the disulfides prevent expression of helicity in regions with alpha-helix potential. Application of the Chou-Fasman secondary structure predictive algorithm to hCG beta suggested that several regions of helix potential, in particular regions 14-21, 59-69, and perhaps 80-88, may account for much of the helicity observed in peptides 1-43 and 44-95, respectively, in TFE. The region from 96-145 has no significant potential for helicity, consistent with the measured circular dichroic spectra of peptides 96-114 and 123-145. These results demonstrate that helicity can occur in the linear form of hCG beta, and this secondary structure can best be attributed to the amino terminal and the middle portion of the molecular. Several potential regions of beta-structure and beta-turns were also suggested.

Amino Acid Sequence↗
PubMed2026

Circular RNAs orchestrate integrated post-transcriptional responses to combined heat and drought stress in rice.

Circular RNAs (circRNAs) are emerging post-transcriptional regulators, yet their landscape and functional roles in rice under combined abiotic stress remain largely unexplored. Here, we systematically reanalyzed strand-specific RNA-seq data to characterize circRNAs responsive to simultaneous heat and drought stress. Following quality control, read mapping, and dual-algorithm prediction using CIRI2 and CIRCexplorer2, we identified 208 high-confidence circRNAs distributed across all 12 chromosomes. Comparative profiling revealed 83 circRNAs uniquely expressed in control samples, 51 in stressed samples, and 74 shared between conditions, indicating stress-dependent circularization. Junction-read analysis highlighted a spectrum of circularization strength, ranging from highly abundant circRNAs with dominant junction reads to low-confidence candidates masked by linear transcript background. Genomic annotation showed that circRNAs primarily originated from exonic and intergenic regions, with a pronounced negative-strand bias; several genes generated multiple circRNA isoforms via alternative back-splicing. Functional enrichment of host genes suggested involvement in protein folding, nutrient reservoir activity, RNA degradation, and branched-chain amino acid catabolism, implicating roles in stress adaptation and metabolic regulation. Differential expression analysis identified seven circRNAs specifically induced under combined stress conditions. Network topology analysis pinpointed key miRNAs-including osa-miR414, osa-miR1439, and osa-miR2919-as candidate topological hubs within the predicted network. Their predicted target genes, such as those encoding stress-responsive transcription factors and signaling proteins, suggest potential roles in coordinating post-transcriptional responses to combined stress. Network topology analysis pinpointed key miRNAs-including osa-miR414, osa-miR1439, and osa-miR2919-as candidate topological hubs within the predicted network. Their predicted target genes, such as those encoding stress-responsive transcription factors and signaling proteins, suggest potential roles in coordinating post-transcriptional responses to combined stress. Overall, this study provides a comprehensive map of circRNAs in rice under combined heat and drought stress, suggests their potential as ceRNAs based on predictive analysis, and lays a foundation for future experimental validation of circRNA-mediated regulation.

Oryza↗
PubMed2026

Machine learning-ready genomic biomarkers: ATF3 polymorphisms predict postoperative analgesic demand through AI-compatible phenotyping.

PURPOSE: To determine whether ATF3 polymorphisms can serve as genetic biomarkers for machine learning-based precision analgesia by establishing a genotype-phenotype association suitable for predictive modeling of postoperative opioid requirements. METHODS: In a prospective cohort of 167 adults undergoing abdominal surgery, ATF3 SNPs rs3122721 and rs3125293 were genotyped. A structured dataset architecture was developed to represent genetic profiles as input features for supervised learning models, enabling translational analysis of genotype‑dependent opioid consumption over 72 h. RESULTS: Patients with homozygous genotypes of the ATF3 SNPs had significantly higher opioid requirements than non‑carriers, despite reporting similar subjective pain scores. This consistent genotype‑dependent pattern provided a clinically relevant phenotype suitable for integration into predictive algorithms. CONCLUSION: ATF3 genotyping offers a promising biomarker for computationally informed precision analgesia. By linking genomic variability to clinically meaningful outcomes within a structured clinical and genomic framework, this approach supports the future development of risk-stratified clinical decision-support systems to optimize postoperative pain management.Trial registration ChiCTR1900021991, registered 30 April 2019. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at https://doi.org/10.1007/s13755-026-00480-9.

ATF3↗
PubMed1992

Immunological relatedness of the sarcoplasmic reticulum Ca(2+)-ATPase and the Na+,K(+)-ATPase.

The effect of anti-ATPase antibodies with epitopes near Asp-351 (PR-8), Lys-515 (PR-11) and the ATP binding domain (D12) of the Ca(2+)-ATPase of sarcoplasmic reticulum (EC 3.6.1.38) was analyzed. The PR-8 and D12 antibodies reacted freely with the Ca(2+)-ATPase in the native membrane, indicating that their epitopes are exposed on the cytoplasmic surface. Both PR-8 and D12 interfered with the crystallization of the Ca(2+)-ATPase, suggesting that their binding sites are at interfaces between ATPase molecules. PR-11 had no effect on ATPase-ATPase interactions or on the ATPase activity of sarcoplasmic reticulum. The epitope of PR-11 is suggested to be the VIDRC sequence at residues 520-525, while that of D12 at residues 670-720 of the Ca(2+)-ATPase. The use of predictive algorithms of antigenicity for identification of potential antigenic determinants in the Ca(2+)-ATPase is analyzed.

Amino Acid Sequence↗
PubMed1989

The surface properties of apolipoproteins A-I and A-II at the lipid/water interface.

The monolayer system was employed to investigate the relative affinities of apolipoproteins A-I and A-II for the lipid/water interface. The adsorption of reductively 14C-methylated apolipoproteins to phospholipid monolayers spread at the air/water interface was determined by monitoring the surface pressure of the mixed monolayer and the surface concentration of the apoprotein. ApoA-II has a higher affinity than apoA-I for lipid monolayers; for a given initial surface pressure, apoA-II adsorbs more than apoA-I to monolayers of egg phosphatidylcholine (PC), distearoyl-PC and human high-density lipoprotein (HDL3) surface lipids. Comparison of the molecular packing of apolipoproteins A-I and A-II suggests that apoA-II adopts a more condensed conformation at the lipid/water interface compared to apoA-I. The ability of apoA-II to displace apoA-I from egg PC and HDL3 surface lipid monolayers was studied by following the adsorption and desorption of the reductively 14C-methylated apolipoproteins. At saturating subphase concentrations of the apoproteins (3.10(-5) g/100 ml), two molecules of apoA-II absorbed for each molecule of apoA-I displaced. This displacement was accompanied by an increase in surface pressure. An identical stoichiometry for the displacement of apoA-I from HDL particles by apoA-II has been reported by others. At low subphase concentrations of apoproteins (5.10(-6) g/100 ml), the apoA-I/lipid monolayer was not fully compressed and could accommodate the adsorbing apoA-II molecules without displacement of apoA-I molecules. ApoA-I molecules were unable to displace apoA-II from the lipid/water interface. The average residue hydrophobicity of apoA-II is higher than that of apoA-I; this may contribute to the higher affinity of apoA-II for lipids compared to apoA-I. The probable helical regions in apolipoproteins A-I and A-II were located using a secondary structure prediction algorithm. The analysis suggests that the amphiphilic properties of the alpha-helical regions of apoA-I and apoA-II are probably not significantly different. Further understanding of the differences in surface activity of these apolipoproteins will require more knowledge of their secondary and tertiary structures.

Amino Acid Sequence↗
PubMed1990

Identification of a potent synthetic HIV1 immunogen compromising gag-P24 tandem T- and B-cell epitopes.

Recent studies indicate that the gag gene products may play a crucial role in the immune response against HIV infection since clinical progression to AIDS is associated with a reduction in the level of circulating antibodies to gag p24 and antibodies raised against p17 peptide can inhibit HIV1 infection in vitro. Using conventional structure prediction algorithms for T-cell and B-cell epitopes, we have selected and chemically synthesized several gag peptides. In particular, an unconjugated HIV1-p24 peptide containing both B- and T-cell epitopes in tandem plus Freund's adjuvant induced a strong antibody response in both mice and rabbits against p24 and its precursor p55 as judged by immunoblotting. In addition, the peptide presented in the appropriate MHC context was shown to be highly stimulatory for p24 specific murine T-cell clones.

Acquired Immunodeficiency Syndrome↗
PubMed1992

Do G protein subunits associate via a three-stranded coiled coil?

We used a computer-based prediction algorithm to identify probable coiled-coil segments at the N-termini of G protein alpha, beta and gamma subunits. This result indicates that G protein trimers may form via a three-stranded coiled coil. Previous biochemical results had shown that the N-termini of alpha and beta are involved in subunit interactions. Here we present a structural model for the N-terminal domain of beta gamma and a hypothesis for the reversible association of alpha to beta gamma.

Algorithms↗
PubMed1992

Definition of linear antigenic regions of the HPV16 L1 capsid protein using synthetic virion-like particles.

Mice of three haplotypes (H-2d, H-2b, and H-2d/b) were immunized with synthetic HPV16 virus-like particles (VLPs), produced using a vaccinia virus doubly recombinant for the L1 and L2 proteins of HPV16. The resultant anti-VLP antisera recognized HPV16 capsids by ELISA assay and baculovirus recombinant HPV16 L1 and L2 protein on immunoblot. Overlapping peptides corresponding to the HPV16 L1 amino acid sequence were used to define the immunoreactive regions of the L1 protein. The majority of the L1 peptides were reactive with IgG from the mice immunized with the synthetic HPV16 capsids. A computer algorithm predicted seven B epitopes in HPV16 L1, five of which lay within peptides strongly reactive with the murine antisera. The murine anti-VLP antisera failed to react with the two peptides recognized by anti-HPV16L1 monoclonal antibodies raised by others against recombinant L1 fusion protein. We conclude that the immunoreactive epitopes of HPV16 defined using virus-like particles differ significantly from those defined using recombinant HPV16 L1 fusion proteins, which implies that such fusion proteins may not be the antigens to look for HPV16L1 specific immune responses in HPV-infected patients.

Amino Acid Sequence↗
PubMed1992

Chromogranin A epitopes: clues from synthetic peptides and peptide mapping.

Chromogranin A (CgA) is a 48 kDa acidic protein in neuroendocrine secretory vesicles whose primary structure is now known. We used synthetic peptides, synthetic peptide antisera, intact molecule antisera, chymotryptic peptide mapping, microsequencing, immunoblotting, and immunoprecipitation to probe the location of immunodominant domains within the CgA molecule. Polyclonal anti mid-molecule, anti N-terminal and anti C-terminal antibodies specifically visualized CgA (both bovine and human) in one and two dimensional immunoblots of adrenal chromaffin vesicles, and the stain CgA fragments further suggested bidirectional (both N- and C-terminal) cleavage or processing of CgA. Anti intact CgA immunoblotting of HPLC-separated peptides from chymotrypsin-digested bovine CgA revealed several strongly immunoreactive internal peptides, two of which were positioned by N-terminal amino acid sequencing: CgA91ff. and CgA197ff.. A single synthetic peptide (CgA79-113) was recognized by three antibodies developed against the intact CgA molecule: two polyclonal rabbit antisera as well as a monoclonal mouse antibody. Not all antigenicity algorithm-predicted domains were immunogenic, suggesting that some of these predicted domains may not be accessible. Polyclonal anti mid-molecule, anti N- and anti C-terminal synthetic peptide antisera specifically immunoprecipitated 125I-labeled bovine CgA from aqueous solution; mid-molecule antisera precipitated substantially greater amounts than terminal antisera. The immunoprecipitation results suggested exposed terminal as well as interior hydrophilic epitopes in the molecule in its intact, native conformation. 125I-human CgA was best precipitated by anti N-terminal antisera, consistent with greatest interspecies sequence conservation at the N-terminus of CgA. The terminal antisera reacted immunohistochemically in a granular pattern with adrenal medullary chromaffin cells (but not adrenal cortical cells) and pancreatic islet cells (but not pancreatic exocrine acini). Thus, synthetic and chymotryptic peptides yielded novel and specific insights into the structure, conformation, vesicular processing and interior immunodominant domains of CgA.

Amino Acid Sequence↗
PubMed1989

Selection of antibody epitopes in an immunopathogenic neural autoantigen.

Results with two well-characterized self-antigens, cytochrome c and myelin basic protein, have led to differing opinions regarding the predominant specificities of autoantibodies, whether regions of sequence diversity or 'structurally inherent features' of a protein determine favored antigenic sites. To further examine this question, 16 antibody epitopes have been mapped on a highly immunopathogenic autoantigen, retinal S-antigen (S-Ag). The epitopes were characterized for: (1) sequence diversity and cross-reactivity on S-antigens from several species; (2) conformational dependency; and (3) probability of their occurrence on the surface of S-antigen. A single C-terminal region containing sequence diversity was most frequently recognized, but no evidence for recognition of any other regions of sequence diversity was found. Thirteen of 16 monoclonal antibodies raised to native S-Ag bound epitopes strongly predicted to be on the surface of S-antigen. Conversely, only one of six antibody preparations raised to peptides or affinity-purified on peptides was found to recognize an epitope predicted to be on the surface, suggesting a good correlation between specificity for conformation-dependent sites and surface probability based on the surface prediction algorithm. Three of these six antibodies which preferred denatured epitopes bound sites which overlapped or coincided with T cell sites; two of these T cell sites are immunopathogenic. The epitopes recognized on denatured antigen and peptides were similar whether the antibodies were elicited with intact human or bovine S-antigen or with cyanogen bromide-cleaved peptides. Our data suggests that in the case of S-antigen, structural features are more significant factors in epitope selection than sequence diversity.

Amino Acid Sequence↗