pith. sign in

arxiv: 2605.21084 · v1 · pith:UEYB77AJnew · submitted 2026-05-20 · ⚛️ physics.bio-ph

Label-free SERS Discrimination of Native Proline Hydroxylation at Single-molecule peptide by Deep Learning-assisted plasmonic nanopore

Yingqi Zhao , Kuo Zhan , Pei-Lin Xin , Yuge Liang , Enock Adjei Agyekum , Matti Putkonen , Shuai Li , Francesco De Angelis
show 1 more author
Jianan Huang
This is my paper

Pith reviewed 2026-05-21 01:38 UTC · model grok-4.3

classification ⚛️ physics.bio-ph
keywords SERSsingle-moleculeproline hydroxylationdeep learningplasmonic nanoporepeptide PTMlabel-free detectionadsorption conformation
0
0 comments X

The pith

A particle-in-pore SERS platform with peak-frequency analysis and a 1D convolutional network discriminates hydroxylated from non-hydroxylated HIF peptides at the single-molecule level.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper sets out to show that native proline hydroxylation, a key post-translational modification, can be read out label-free from individual short peptides by combining surface-enhanced Raman scattering inside a nanopore with statistical peak analysis and neural-network classification. It tests three peptide lengths centered on the HIF Pro-564 site and reports that spectral patterns tied to adsorption conformation become more distinguishable as chain length increases. A reader would care because most current methods either average over ensembles or require chemical tags, leaving single-molecule PTM detection largely inaccessible. The work demonstrates that the combined physical and computational pipeline produces classification accuracies that rise from roughly 73 percent for the shortest pair to nearly 90 percent for the longest, with area-under-curve values above 0.80 in all cases.

Core claim

The central claim is that the particle-in-pore SERS sensor, when paired with peak occurrence frequency analysis and a one-dimensional convolutional neural network, captures reproducible hydroxylation-dependent changes in adsorption conformation and surface interactions, enabling reliable single-molecule discrimination of hydroxylated versus non-hydroxylated HIF peptide fragments. For the 7-, 9-, and 15-amino-acid pairs the post-evaluation accuracies reach 72.98 percent, 78.55 percent, and 89.74 percent respectively, each with AUC above 0.80. Gradient-weighted visualization shows that the network attends to the same recurrent spectral features identified by the peak-frequency method, and an增强

What carries the argument

The particle-in-pore single-molecule SERS platform that records Raman spectra while peptides adsorb to gold nanoparticles inside a nanopore, processed by peak occurrence frequency statistics and 1D-CNN classification.

If this is right

  • Hydroxylation alters how longer peptides adsorb to the gold surface, amplifying spectral differences and raising classification accuracy.
  • The same platform can register weak PTM signatures without labels or ensemble averaging.
  • Gradient-weighted maps confirm that the neural network learns chemically meaningful features that match the peak-occurrence statistics.
  • The approach works across peptide lengths from 7 to 15 amino acids, suggesting scalability to other HIF-derived fragments.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the adsorption change proves general for other polar PTMs, the same hardware could be used to screen multiple modification types without redesign.
  • Integration with microfluidic delivery might allow the sensor to monitor dynamic hydroxylation states in real time rather than in static snapshots.
  • The citrate-band enhancement observed for the longest peptide hints that surface-chemistry tuning could further boost contrast for shorter or more rigid sequences.

Load-bearing premise

The observed spectral and adsorption differences are produced specifically by the presence or absence of the hydroxyl group on proline rather than by uncontrolled differences in peptide conformation, surface coverage, or citrate binding.

What would settle it

If controlled experiments that fix peptide conformation while removing the hydroxylation produce identical SERS spectra and identical citrate-band intensities, the claim that hydroxylation itself drives the distinguishable signals would be falsified.

Figures

Figures reproduced from arXiv: 2605.21084 by Enock Adjei Agyekum, Francesco De Angelis, Jianan Huang, Kuo Zhan, Matti Putkonen, Pei-Lin Xin, Shuai Li, Yingqi Zhao, Yuge Liang.

Figure 2
Figure 2. Figure 2: (a)Distribution of the normalized POF of HIF7S, HIF7SPTM, HIF9S, and HIF9SPTM, shown in red, orange, dark green, and light green, respectively. The peaks discussed in the text are highlighted by yellow and gray bands and labelled with their corresponding Raman shifts. Yellow-shaded bands indicate the region with similar spectra changes for 7AA and 9AA peptide pairs and can be tentatively assigned, whereas … view at source ↗
Figure 5
Figure 5. Figure 5: (a) Normalized POF profiles of HIF15S, HIF15SPTM, and pure AuNPs, shown in dark blue, light blue, and magenta, respectively. The blue-shaded bands indicate peptide-related peaks, while the red-shaded bands highlight characteristic citrate peaks. The offset of the HIF15SPTM POF curve is 0.05. (b) Proposed peptide adsorption states on citrate-covered AuNPs. The lower panel illustrates citrate-covered AuNPs w… view at source ↗
read the original abstract

Post-translational modifications (PTMs) play essential roles in regulating protein structure, function, and cellular signalling. However, peptide level discrimination of hydroxylation at the single-molecule level remains difficult. Here, we report a particle-in-pore single-molecule surface-enhanced Raman spectroscopy (SERS) platform combined with peak occurrence frequency (POF) analysis and a one-dimensional convolutional neural network (1D-CNN) for discriminating hydroxylated and non-hydroxylated HIF peptide fragments. Three peptide pairs containing the Pro-564 hydroxylation site, with lengths of 7, 9, and 15 amino acids (AAs), were investigated. POF analysis revealed reproducible hydroxylation-dependent spectral changes in the 7AA and 9AA peptide pairs, which were attributed to changes in adsorption conformation and surface interactions. CNN-based classification achieved post-evaluation accuracies of 72.98%, 78.55%, and 89.74% for the 7AA, 9AA, and 15AA peptide pairs, respectively, with AUC values above 0.80 for all the pairs, indicating a reliable discrimination. Gradient-weighted feature visualization further showed that CNN-sensitive regions overlapped with recurrent POF features, supporting the chemical relevance of the learned classification patterns. Notably, for the 15AA peptide pair, the enhanced citrate-associated band suggests that hydroxylation can substantially alter peptide-gold nanoparticle adsorption behaviour. This adsorption-mediated effect may amplify hydroxylation-induced spectral differences and contribute to the improved discrimination accuracy despite the increased structural complexity. These results demonstrate that the particle-in-pore sensor, assisted by deep learning, can capture hydroxylation-induced spectral and adsorption changes in peptide fragments, providing a promising strategy for ultrasensitive analysis of weak PTM signatures in peptides.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript presents a particle-in-pore single-molecule SERS platform integrated with peak occurrence frequency (POF) analysis and a 1D-CNN for label-free discrimination of hydroxylated versus non-hydroxylated HIF peptide fragments at the Pro-564 site. Three peptide length pairs (7AA, 9AA, 15AA) are examined; POF identifies reproducible spectral changes attributed to adsorption conformation shifts, and the CNN achieves post-evaluation accuracies of 72.98%, 78.55%, and 89.74% with AUC values >0.80. Gradient-weighted visualizations link CNN features to POF peaks, and an enhanced citrate band is noted for the 15AA pair as evidence of hydroxylation-altered AuNP adsorption.

Significance. If the spectral and classification differences prove specific to proline hydroxylation rather than uncontrolled adsorption or conformational variations, the work would offer a promising route for ultrasensitive, label-free PTM detection in peptides where conventional methods lack sensitivity. The combination of experimental SERS nanopore measurements, POF feature extraction, CNN classification with AUC>0.80, and feature visualization provides concrete empirical support and chemical interpretability that strengthens the central claim.

major comments (2)
  1. [Abstract] Abstract: The central claim that the platform discriminates native proline hydroxylation requires that observed spectral shifts and CNN accuracies arise specifically from the PTM. The abstract itself states that for the 15AA pair 'the enhanced citrate-associated band suggests that hydroxylation can substantially alter peptide-gold nanoparticle adsorption behaviour' and that this 'may amplify hydroxylation-induced spectral differences.' No description is given of matched controls that hold peptide concentration, ionic strength, or conformational ensemble fixed while toggling only the hydroxylation state; without such controls the CNN decision boundaries could be driven by surface-coverage or citrate-interaction differences instead of the PTM.
  2. [Abstract] Abstract and POF analysis section: Reproducibility of hydroxylation-dependent spectral changes is asserted, yet the manuscript provides no quantitative metrics (e.g., standard deviations on peak frequencies or intensities across independent measurements) or explicit statement of how many independent peptide preparations and nanopore runs were performed for each pair. This information is load-bearing for the claim that the 72.98–89.74% accuracies reflect hydroxylation rather than batch-to-batch variability.
minor comments (2)
  1. [Abstract] The abstract reports concrete accuracies and AUC thresholds but omits raw spectra, training/validation splits, cross-validation procedure, or error bars on the reported accuracies.
  2. Notation for the 1D-CNN architecture (layer counts, kernel sizes, training hyperparameters) is not summarized in the abstract or methods overview, making it difficult to assess reproducibility of the 72.98–89.74% figures.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We are grateful to the referee for the thorough review and valuable suggestions. We address the major comments point by point below, indicating where revisions will be made to improve the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central claim that the platform discriminates native proline hydroxylation requires that observed spectral shifts and CNN accuracies arise specifically from the PTM. The abstract itself states that for the 15AA pair 'the enhanced citrate-associated band suggests that hydroxylation can substantially alter peptide-gold nanoparticle adsorption behaviour' and that this 'may amplify hydroxylation-induced spectral differences.' No description is given of matched controls that hold peptide concentration, ionic strength, or conformational ensemble fixed while toggling only the hydroxylation state; without such controls the CNN decision boundaries could be driven by surface-coverage or citrate-interaction differences instead of the PTM.

    Authors: We appreciate this concern regarding the specificity to the PTM. The hydroxylated and non-hydroxylated peptides differ only at the Pro-564 site, with all other experimental conditions, including concentration and buffer, kept identical for each pair. The observed differences in POF and the CNN's reliance on specific spectral regions (as shown by gradient-weighted visualizations) support that the discrimination is due to hydroxylation-induced changes in adsorption and conformation. However, to address the referee's point directly, we will revise the abstract and methods to explicitly describe the matched controls and provide additional justification that citrate effects are a consequence of the PTM rather than a confounding factor. We will also discuss potential limitations in controlling conformational ensembles at the single-molecule level. revision: yes

  2. Referee: [Abstract] Abstract and POF analysis section: Reproducibility of hydroxylation-dependent spectral changes is asserted, yet the manuscript provides no quantitative metrics (e.g., standard deviations on peak frequencies or intensities across independent measurements) or explicit statement of how many independent peptide preparations and nanopore runs were performed for each pair. This information is load-bearing for the claim that the 72.98–89.74% accuracies reflect hydroxylation rather than batch-to-batch variability.

    Authors: We concur that providing quantitative metrics and replicate information is crucial. The experiments were conducted with multiple independent preparations and runs to ensure reproducibility, but these details were omitted from the main text for brevity. In the revised manuscript, we will include a dedicated subsection on experimental reproducibility, reporting the number of independent preparations (n ≥ 3 for each peptide pair) and total measurements, along with standard deviations for the key POF peak positions and intensities. This will substantiate that the classification accuracies are not due to batch variability. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical spectra + independent CNN training

full rationale

The paper presents an experimental workflow: measured SERS spectra from hydroxylated vs non-hydroxylated peptide samples, POF feature extraction on those spectra, and standard 1D-CNN training/evaluation yielding reported accuracies (72.98–89.74 %, AUC > 0.80). No equations, fitted parameters, or self-citations are invoked such that any output reduces to the input by construction. The classification result is obtained from held-out evaluation on independently acquired spectral data; the interpretation that differences arise from hydroxylation is an empirical claim open to controls, not a definitional or self-referential loop. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The platform rests on standard SERS plasmonic enhancement and the assumption that hydroxylation alters adsorption conformation; no explicit free parameters or new entities are introduced in the abstract.

axioms (2)
  • domain assumption Plasmonic enhancement in the particle-in-pore geometry produces measurable SERS spectra from single peptides
    Invoked in the platform description to justify single-molecule sensitivity.
  • domain assumption Hydroxylation-induced changes in peptide adsorption produce reproducible shifts in peak occurrence frequencies
    Central to attributing POF differences to the PTM rather than experimental noise.

pith-pipeline@v0.9.0 · 5891 in / 1396 out tokens · 36634 ms · 2026-05-21T01:38:00.117123+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    POF analysis revealed reproducible hydroxylation-dependent spectral changes... attributed to changes in adsorption conformation and surface interactions. CNN-based classification achieved post-evaluation accuracies of 72.98%, 78.55%, and 89.74%

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

56 extracted references · 56 canonical work pages

  1. [1]

    Frontiers in Oncology, 2017

    Iommarini, L., et al., Non-Canonical Mechanisms Regulating Hypoxia-Inducible Factor 1 Alpha in Cancer. Frontiers in Oncology, 2017. 7

    work page 2017

  2. [2]

    Han, and B

    Ma, H., X.X. Han, and B. Zhao, Enhanced Raman spectroscopic analysis of protein post- translational modifications. TrAC Trends in Analytical Chemistry, 2020. 131: p. 116019

    work page 2020

  3. [3]

    Trends in Pharmacological Sciences, 2012

    Semenza, G.L., Hypoxia-inducible factors: mediators of cancer progression and targets for cancer therapy. Trends in Pharmacological Sciences, 2012. 33(4): p. 207-214

    work page 2012

  4. [4]

    International Journal of Analytical Chemistry, 2016, 7436849, 6 pages, 2016

    Ponomarenko, E., et al., The Size of the Human Proteome: The Width and Depth. International Journal of Analytical Chemistry, 2016, 7436849, 6 pages, 2016

    work page 2016

  5. [5]

    Angewandte Chemie-International Edition, 2025

    Li, Z., et al., Nanopore-Based High-Resolution Detection of Multiple Post-Translational Modifications in Protein. Angewandte Chemie-International Edition, 2025. 64(15)

    work page 2025

  6. [6]

    Rodriguez-Larrea, and H

    Rosen, C., D. Rodriguez-Larrea, and H. Bayley, Single-molecule site-specific detection of protein phosphorylation with a nanopore. Nature Biotechnology, 2014. 32(2): p. 179-181

    work page 2014

  7. [7]

    J Am Chem Soc, 2022

    Ensslen, T., et al., Resolving Isomeric Posttranslational Modifications Using a Biological Nanopore as a Sensor of Molecular Shape. J Am Chem Soc, 2022. 144(35): p. 16060-16068

    work page 2022

  8. [8]

    Biomaterials, 2021

    Zhang, L., et al., Detection of single peptide with only one amino acid modification via electronic fingerprinting using reengineered durable channel of Phi29 DNA packaging motor. Biomaterials, 2021. 276: p. 121022

    work page 2021

  9. [9]

    Fahie, M.A. and M. Chen, Electrostatic Interactions between OmpG Nanopore and Analyte Protein Surface Can Distinguish between Glycosylated Isoforms. J Phys Chem B, 2015. 119(32): p. 10198-206

    work page 2015

  10. [10]

    ACS Nano 2024, 18, 2, 1504–1515

    Chan Cao , et al., Deep learning-assisted single-molecule detection of protein post- translational modifications with a biological nanopore. ACS Nano 2024, 18, 2, 1504–1515

    work page 2024

  11. [11]

    Angewandte Chemie-International Edition, 2025

    Chang, J., et al., Precise Identification of Native Peptides with Posttranslational Proline Hydroxylation by Nanopore. Angewandte Chemie-International Edition, 2025. 64(9)

    work page 2025

  12. [12]

    Analytical Chemistry, 2006

    Sundararajan, N., et al., Ultrasensitive detection and characterization of posttranslational modifications using surface-enhanced Raman spectroscopy. Analytical Chemistry, 2006. 78(11): p. 3543-3550

    work page 2006

  13. [13]

    Advanced Functional Materials,

    Iarossi, M., et al., High-Density Plasmonic Nanopores for DNA Sensing at Ultra-Low Concentrations by Plasmon-Enhanced Raman Spectroscopy. Advanced Functional Materials,

    work page

  14. [14]

    Nature Communications, 2025

    Khabarov, K., et al., Raman identification of single nucleotides flowing through permeable plasmonic films. Nature Communications, 2025. 16(1): p. 9113

    work page 2025

  15. [15]

    Crystengcomm, 2018

    Du, J., et al., The role of phosphorylation and dephosphorylation of shell matrix proteins in shell formation: an in vivo and in vitro study. Crystengcomm, 2018. 20(27): p. 3905-3916

    work page 2018

  16. [16]

    Analytical Chemistry, 2024, 96, 45, 17978–17983

    Geddis, A., et al., Label-Free SERS Sensors for Real-Time Monitoring of Tyrosine Phosphorylation. Analytical Chemistry, 2024, 96, 45, 17978–17983

    work page 2024

  17. [17]

    Biosensors & Bioelectronics, 2015

    Li, Y ., et al., Label-free in situ monitoring of protein tyrosine nitration in blood by surface- enhanced Raman spectroscopy. Biosensors & Bioelectronics, 2015. 69: p. 1-7

    work page 2015

  18. [18]

    Journal of Photochemistry and Photobiology B- Biology, 2018

    Sjöberg, B., et al., Methionine oxidation by hydrogen peroxide in peptides and proteins: A theoretical and Raman spectroscopy study. Journal of Photochemistry and Photobiology B- Biology, 2018. 188: p. 95-99

    work page 2018

  19. [19]

    Guerrero, A. and R. Aroca, Surface-enhanced Raman scattering of hydroxyproline. Journal of Raman Spectroscopy, 2012. 43(4): p. 478-481

    work page 2012

  20. [20]

    A Raman, SERS and theoretical study

    Cárcamo, J., et al., Proline and hydroxyproline deposited on silver nanoparticles. A Raman, SERS and theoretical study. Journal of Raman Spectroscopy, 2012. 43(6): p. 750-755

    work page 2012

  21. [21]

    Chemical Society Reviews, 2019

    Lee, H., et al., Designing surface-enhanced Raman scattering (SERS) platforms beyond hotspot engineering: emerging opportunities in analyte manipulations and hybrid materials. Chemical Society Reviews, 2019. 48(3): p. 731-756

    work page 2019

  22. [22]

    Angewandte Chemie International Edition, 2020

    Huang, J.-A., et al., Multiplexed Discrimination of Single Amino Acid Residues in Polypeptides in a Single SERS Hot Spot. Angewandte Chemie International Edition, 2020. 59(28): p. 11423- 11431

    work page 2020

  23. [23]

    Nature Communications, 2019

    Huang, J.-A., et al., SERS discrimination of single DNA bases in single oligonucleotides by electro-plasmonic trapping. Nature Communications, 2019. 10(1): p. 5321

    work page 2019

  24. [24]

    Acs Photonics, 2022

    Zhao, Y ., et al., Label-Free Optical Analysis of Biomolecules in Solid-State Nanopores: Toward Single-Molecule Protein Sequencing. Acs Photonics, 2022. 9(3): p. 730-742

    work page 2022

  25. [25]

    Angewandte Chemie- International Edition, 2025

    Lim, S., et al., Unlocking the Full Potential of SERS: Merging Direct and Indirect Approaches for Enhanced Analysis of Multiplex Plasticizer Analogs in Matrices. Angewandte Chemie- International Edition, 2025. 64(40)

    work page 2025

  26. [26]

    Angewandte Chemie-International Edition, 2026

    Tang, W., et al., Digital Decoding of Multicomponent Protein Systems via Nanocavity-Confined Single-Molecule Raman Fingerprinting. Angewandte Chemie-International Edition, 2026. 65(6)

    work page 2026

  27. [27]

    Acs Photonics, 2026

    Khabarov, K., et al., Single-Molecule Peptide Discrimination via Flow-Through SERS and Machine Learning. Acs Photonics, 2026. 13(7): p. 1840-1845

    work page 2026

  28. [28]

    Nano Letters, 2025, 25, 18, 7499–7506

    Zhao, Y ., et al., Single-Molecule SERS Discrimination of Proline from Hydroxyproline Assisted by a Deep Learning Model. Nano Letters, 2025, 25, 18, 7499–7506

    work page 2025

  29. [29]

    Abramson, J., et al., Accurate structure prediction of biomolecular interactions with AlphaFold

    work page

  30. [30]

    630(8016): p

    Nature, 2024. 630(8016): p. 493-500

    work page 2024

  31. [32]

    Nature Communications, 2020

    Franco-Ulloa, S., et al., Dispersion state phase diagram of citrate-coated metallic nanoparticles in saline solutions. Nature Communications, 2020. 11(1)

    work page 2020

  32. [33]

    Buglak, A. and A. Kononov, Comparative study of gold and silver interactions with amino acids and nucleobases. Rsc Advances, 2020. 10(56): p. 34149-34160

    work page 2020

  33. [34]

    Chemical Reviews, 2017

    Zrimsek, A., et al., Single-Molecule Chemistry with Surface- and Tip-Enhanced Raman Spectroscopy. Chemical Reviews, 2017. 117(11): p. 7583-7613

    work page 2017

  34. [35]

    Nano Letters, 2018

    Shin, H., et al., Frequency-Domain Proof of the Existence of Atomic-Scale SERS Hot-Spots. Nano Letters, 2018. 18(1): p. 262-271

    work page 2018

  35. [36]

    Acs Nano, 2024

    Li, W., et al., High-Throughput Single-Molecule Surface-Enhanced Raman Spectroscopic Profiling of Single-Amino Acid Substitutions in Peptides by a Gold Plasmonic Nanopore. Acs Nano, 2024. 18(29): p. 19200-19207

    work page 2024

  36. [37]

    Graff, M. and J. Bukowska, Surface-enhanced Raman scattering (SERS) spectroscopy of enantiomeric and racemic methionine on a silver electrode-evidence for chiral discrimination in interactions between adsorbed molecules. Chemical Physics Letters, 2011. 509(1-3): p. 58- 61

    work page 2011

  37. [38]

    Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 2011

    Zhu, G., et al., Raman spectra of amino acids and their aqueous solutions. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 2011. 78(3): p. 1187-1195

    work page 2011

  38. [39]

    Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 2024, 308, 123681

    Hernández, B., et al., The relationship between the tyrosine residue 850-830 cm-1Raman doublet intensity ratio and the aromatic side chain x1 torsion angle. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 2024, 308, 123681

    work page 2024

  39. [40]

    Wolpert, M. and P . Hellwig, Infrared spectra and molar absorption coefficients of the 20 alpha amino acids in aqueous solutions in the spectral range from 1800 to 500 cm-1. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 2006. 64(4): p. 987-1001

    work page 2006

  40. [41]

    Journal of Molecular Liquids, 2025

    Smirnov, M., et al., Interaction of gold and platinum nanoparticles with L-tyrosine in aqueous solution: Conformational and dynamic changes. Journal of Molecular Liquids, 2025. 420: p. 126822

    work page 2025

  41. [42]

    Stewart, S. and P . Fredericks, Surface-enhanced Raman spectroscopy of amino acids adsorbed on an electrochemically prepared silver surface. Spectrochimica Acta Part a- Molecular and Biomolecular Spectroscopy, 1999. 55(7-8): p. 1641-1660

    work page 1999

  42. [43]

    Hernández, B., et al., Vibrational Analysis of Amino Acids and Short Peptides in Hydrated Media. IV . Amino Acids with Hydrophobic Side Chains: L-Alanine, L-Valine, and L-Isoleucine. Journal of Physical Chemistry B, 2009. 113(10): p. 3169-3178

    work page 2009

  43. [44]

    Stewart, S. and P . Fredericks, Surface-enhanced Raman spectroscopy of peptides and proteins adsorbed on an electrochemically prepared silver surface. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 1999. 55(7-8): p. 1615-1640

    work page 1999

  44. [45]

    and T .V

    Herlinger, A.W. and T .V . Long, II, Laser-Raman and infrared spectra of amino acids and their metal complexes. III. Proline and bisprolinato complexes. Journal of the American Chemical Society, 1970. 92(22): p. 6481-6486

    work page 1970

  45. [46]

    Journal of Physical Chemistry, 1983

    SIIMAN, O., et al., SURFACE-ENHANCED RAMAN-SCATTERING BY CITRATE ON COLLOIDAL SILVER. Journal of Physical Chemistry, 1983. 87(6): p. 1014-1023

    work page 1983

  46. [47]

    Langmuir, 1995

    MUNRO, C., et al., CHARACTERIZATION OF THE SURFACE OF A CITRATE-REDUCED COLLOID OPTIMIZED FOR USE AS A SUBSTRATE FOR SURFACE-ENHANCED RESONANCE RAMAN- SCATTERING. Langmuir, 1995. 11(10): p. 3712-3720

    work page 1995

  47. [48]

    Analyst, 2013

    Kurouski, D., et al., Amide I vibrational mode suppression in surface (SERS) and tip (TERS) enhanced Raman spectra of protein specimens. Analyst, 2013. 138(6): p. 1665-1673

    work page 2013

  48. [49]

    Rauls, and V

    Deckert-Gaudig, T., E. Rauls, and V . Deckert, Aromatic Amino Acid Monolayers Sandwiched between Gold and Silver: A Combined Tip-Enhanced Raman and Theoretical Approach. Journal of Physical Chemistry C, 2010. 114(16): p. 7412-7420

    work page 2010

  49. [50]

    Hernández, B., et al., Vibrational Analysis of Amino Acids and Short Peptides in Hydrated Media. VIII. Amino Acids with Aromatic Side Chains: L-Phenylalanine, L-Tyrosine, and L- Tryptophan. Journal of Physical Chemistry B, 2010. 114(46): p. 15319-15330

    work page 2010

  50. [51]

    Journal of Raman Spectroscopy, 2022

    Li, J., et al., Machine learning using convolutional neural networks for SERS analysis of biomarkers in medical diagnostics. Journal of Raman Spectroscopy, 2022. 53(12): p. 2044- 2057

    work page 2022

  51. [52]

    Analytical Methods, 2023

    Lu, P ., et al., CNN-assisted SERS enables ultra-sensitive and simultaneous detection of Scr and BUN for rapid kidney function assessment. Analytical Methods, 2023. 15(3): p. 322-332

    work page 2023

  52. [53]

    Analytical Chemistry,

    Shi, G., et al., 1D Gradient-Weighted Class Activation Mapping, Visualizing Decision Process of Convolutional Neural Network-Based Models in Spectroscopy Analysis. Analytical Chemistry,

    work page

  53. [54]

    9959-9966

    95(26): p. 9959-9966

    work page

  54. [55]

    Analytical and Bioanalytical Chemistry, 2023

    Xiong, C., et al., Rapid and precise detection of cancers via label-free SERS and deep learning. Analytical and Bioanalytical Chemistry, 2023. 415(17): p. 3449-3462

    work page 2023

  55. [56]

    Plos One,

    Tien, M., et al., Maximum Allowed Solvent Accessibilites of Residues in Proteins. Plos One,

    work page

  56. [57]

    8(11): p. e80635. Supporting Information Label-free SERS Discrimination of Native Proline Hydroxylation at Single-molecule peptide by Deep Learning-assisted plasmonic nanopore Yingqi Zhao1,3 Kuo Zhan 1,3, Pei-Lin Xin1,3, Yuge Liang1,3, Enock Adjei Agyekum4, Matti Putkonen5, Shuai Li4, Francesco De Angelis6 and Jianan Huang 1,2,3,* 1 Research Unit of Healt...

    work page 2024