Pith. sign in

REVIEW 4 major objections 6 minor 32 references

SENSAAS (SENsitive Surface As A Shape): utilizing open-source algorithms for 3D point cloud alignment of molecules

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Point-cloud representations of molecular surfaces can be aligned to reveal shared substructures and bioisosteric groups.

desk verdict A plausible open-source alternative to commercial shape alignment, but the evidence is too thin: no code, no baseline, and scores that were tuned on the same cases used for validation. read the letter →

arxiv 1908.11267 v1 pith:QKJEU3XW submitted 2019-08-19 q-bio.BM eess.SP

Dominique Douguet (IPMC , UCA) , Frédéric Payan (UCA) This is my paper · ORCID
classification q-bio.BMeess.SP
keywords molecularsimilarityshapealignmentpointcloudregistrationvanderWaalssurfaceFPFHdescriptorscoloredICPbioisosterismscaffoldhopping
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to show that molecular shape comparison can be treated as a point-cloud registration problem and solved with open-source algorithms originally built for computer vision. A molecule is represented by points spread over its van der Waals surface, each colored by one of four physicochemical atom classes, and two molecules are aligned by first finding a rough global match from geometric descriptors and then refining it with a color-and-geometry-aware registration. The payoff, if the claim is right, is a simple and deployable way to compare small molecules, peptides, proteins, and even binding cavities as negative surfaces, with direct use in scaffold hopping. The paper's substantive claim is that this pipeline identifies and aligns similar shapes and sub-shapes even when the starting point clouds differ significantly, as with bioisosteric tetrazole and carboxylate groups.

What carries the argument

The central object is a colored point cloud: points placed at roughly 0.3-unit spacing on the van der Waals surface, each colored according to four user-defined physicochemical classes (apolar hydrogens and halogens; polar and hydrogen-bonding atoms including polar hydrogens and fluorine; carbon, phosphorus, and boron skeleton atoms; and all other atoms). The argument is carried by a two-step registration pipeline. Global registration computes FPFH histograms for down-sampled points, matches those histograms with RANSAC to propose an initial rotation and translation, and then colored point-cloud registration refines the alignment by iteratively minimizing a functional that mixes three-dimensional geometry and RGB color information. To avoid committing to a single down-sampling scale, the whole procedure is run for eleven voxel sizes from 0.2 to 1.2 and the best COLOR gfit score is retained. The evaluation scores gfit, cfit, and hfit are Tversky-like coefficients measuring, respectively, overall matched points, matched points per color class, and matched points excluding the apolar class.

What would settle it

Run SENSAAS on a benchmark of known bioisosteric pairs in diverse scaffolds and compare the best COLOR gfit pose with the pose that maximizes hfit; if the top-gfit alignment frequently places the two bioisosteric groups in different regions, or if high hfit poses are systematically ranked below low-hfit poses, then the colored point-cloud representation and the best-gfit selection rule are not sufficient for the claimed bioisosteric alignment.

Watch

Extended reading notes

Core claim

The central claim is that SENSAAS, a workflow combining global feature-based registration (Fast Point Feature Histograms matched by RANSAC) with colored point-cloud registration, produces good superimpositions of molecular structures and substructures even when the two point clouds are substantially different. The supporting observations include exact recovery of rigidly moved molecules, correct placement of three imatinib fragments on imatinib, alignment of a tetrazole group onto the carboxylate and carboxylic-acid forms of Adapalene, and alignment of a shared substructure between tranylcypromine and milnacipran. In the Adapalene/tetrazole case, the paper shows that the hfit score, which ignores the apolar class, picks out the chemically meaningful bioisosteric alignment over a geometrically similar but chemically wrong alternative. The stated conclusion is that the method performs well in aligning drug-sized molecules globally and in aligning substructures, fragments, and bioisosteric groups locally.

Load-bearing premise

The load-bearing premise is that a van der Waals surface sampled at roughly 0.3 spacing and colored by four broad atom classes retains enough geometric and physicochemical information for FPFH-based global alignment and colored refinement to recover chemically correct matches, even for small fragments and substructures.

Editorial extensions

If this is right

  • Substructure and fragment matching becomes feasible without a precomputed common core: a small fragment's point cloud is dropped onto a larger molecule's surface and finds its matching sub-shape.
  • Bioisosteric replacements such as tetrazole for carboxylate can be superimposed, which gives scaffold hopping a surface-based criterion for when two chemically different groups occupy the same shape and polar region.
  • The score distribution from 500,000 drug-pair alignments offers a practical screening threshold: roughly 7 percent of pairs score above 0.5 and 2 percent above 0.6, so high COLOR gfit values are selective for similarity.
  • Because the representation is not tied to a particular molecule type, the same workflow can be applied to peptides, proteins, or cavities described as negative images of protein surfaces.
  • The hfit score provides a chemical check on alignment quality, allowing a geometrically plausible pose to be rejected when its polar and aromatic points are not matched.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension would be to enrich the four color classes with computed properties such as partial charges; the paper's own prediction is that this creates many small color patches and makes refinement harder, so the four-class setting is a deliberate trade-off rather than the best possible encoding.
  • The current selection rule picks the alignment with the best COLOR gfit, yet the Adapalene/tetrazole example shows that a slightly lower-gfit pose can be chemically correct; a combined gfit-cfit-hfit selection criterion might change which pose is reported.
  • Since the evaluation threshold of 0.3 is tied to the native point spacing, down-sampled clouds may produce different score scales; a density-aware threshold could alter the proposed screening cutoffs.
  • The method is explicitly conformation-dependent, so a practical virtual-screening protocol would need to align ensembles of conformers, which scales the computational cost and tests whether the seconds-per-alignment speed survives millions of comparisons.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper presents SENSAAS, a workflow that aligns and compares molecular shapes by treating van der Waals surfaces as colored 3D point clouds. The pipeline generates point clouds with the nsc program, assigns each point one of four atom-class colors, performs initial alignment with Open3D's FPFH-based global registration, refines with the colored ICP method, and reports three Tversky-like scores (gfit, cfit, hfit). The authors optimize voxel sizes and RANSAC/ICP parameters, then evaluate on test cases including moved molecules, substructures, tetrazole/carboxylate bioisosterism, dissimilar drug pairs, and conformers. The central claim, stated in the Discussion, is that SENSAAS can identify and align similar shapes and sub-shapes, producing good superimpositions even when point clouds differ significantly.

Significance. If the central claim were established, SENSAAS would be a useful open-source tool for molecular shape alignment, fragment matching, and scaffold hopping. The paper has genuine strengths: it builds entirely on open-source components (Open3D, nsc, BioPandas), the parameter choices are transparently tabulated (Table 1), and the idea of coloring vdW surface points by four physicochemical classes is simple and intuitive. The reported test cases cover several practically relevant scenarios. However, the current evidence is preliminary: the evaluation relies on qualitative visual inspection and the method's own matching scores, parameters were tuned on the same test cases later reported as successes, and no external baseline or null distribution is provided. The paper itself acknowledges that a broader comparison to other tools is needed and that a more detailed analysis of the scores will be performed in the future.

major comments (4)
  1. [Optimization of the voxel size; Eq. (8)] The reported COLOR gfit scores are maxima over eleven down-sampled runs (voxel sizes 0.2 to 1.2), and gfit is normalized by the Source point count only. Consequently, a small fragment placed anywhere inside a larger target can achieve a high gfit by geometric containment, and the best-of-11 protocol induces selection bias. The paper itself demonstrates this ambiguity: in Fig. 6i the wrong tetrazole pose scores gfit 0.612 versus 0.603 for the correct pose, with only hfit (0.041 versus 0.584) discriminating. Without a null distribution of gfit for random fragment placements, the reported fragment gfit values do not establish chemically correct alignment.
  2. [Pairwise alignments; Discussion] The validation relies on visual inspection and the method's own scores; no comparison to an external shape-alignment tool is provided, even though ROCS ComboScores are quoted for the Adapalene/Irbesartan/Valsartan pairs in the same section. A baseline such as random-placement superposition or an established overlay program would be needed to interpret the reported gfit values. The Discussion itself states that 'a broader study would be important to assess performance when compared to other tools,' which underscores that the central claim is not yet supported by comparative evidence.
  3. [Figure 4; Molecular set a) parameter setting] The voxel-size grid and the GLOBAL/COLOR parameters were chosen by observing the same Sorbate, Imatinib, and Imatinib-part2 cases that are later reported as successes (the text says 'Molecular set a) was used to set parameters'). The conclusion that no single voxel size works and that all eleven should be run with the best gfit selected therefore risks overfitting to these test cases. A prospective evaluation on held-out cases, or at least a cross-validated parameter-selection procedure, is required to support the claim that the method 'performs well' generally.
  4. [Equations (8)-(10); Results on 500,000 pairs] The scores gfit, cfit, and hfit are described as discriminating similar from dissimilar molecules, and a gfit threshold of 0.5 is suggested from the 500,000-pair distribution. However, this distribution is called 'preliminary' and the scores are computed after alignment selected by gfit alone; hfit is not used to choose among alternative poses (in Fig. 6, hfit is inspected post hoc). The claim that gfit > 0.5 indicates 'clear similarity with reproducible results' is therefore not yet a validated decision rule.
minor comments (6)
  1. [Pairwise alignments] Sorbate/SorbaceC appears in the text and should be Sorbate/SorbateC.
  2. [Figure 7 legend] The legend contains the typo 'Adapalen' instead of 'Adapalene'.
  3. [Figure 4 caption] The caption says three runs are plotted (blue, orange, and mauve lines), but the figure panels as described appear to show only one or two curves; please clarify what each line and marker represents.
  4. [Equations] Equation numbering starts at (8), while Eqs. (1)-(7) are not used; please renumber or remove the unused references.
  5. [Discussion] The word 'librairies' is a typo for 'libraries'.
  6. [Methods I-1] The manuscript compares with Baum et al. in the Discussion, but it does not state why the vdW surface is used instead of the solvent-excluded surface; a sentence clarifying this choice would help readers interpret the point-cloud resolution claims.

Circularity Check

2 steps flagged · score 5.0 of 10

Reported success scores are partly constructed: voxel-grid selection maximizes COLOR gfit on the same cases later reported as validation, and no external or null benchmark anchors the gfit threshold.

  1. fitted input called prediction [Results, 'Optimization of the voxel size' and 'Pairwise alignments'; Methods, Fig. 2 caption]
    ""Molecular set a) was used to set parameters and successful superimpositions were obtained. Molecular set b) was used to investigate substructure matchings." "we choose to execute the GLOBAL and COLOR methods with each voxel size ranging from 0.2 to 1.2 with an increment of 0.1. ... The best COLOR gfit score is selected as the best alignment for the two point clouds.""

    The pairs whose alignments are reported as evidence of success (Sorbate/Sorbate(moved), Imatinib/Imatinib(moved), Imatinib/Imatinib-part2) are the same pairs used in Figure 4 to choose the multi-voxel protocol. The protocol then reports the maximum COLOR gfit over eleven voxel sizes, so the displayed score is the optimum of the selection rule rather than an independent measure. The central claim that SENSAAS 'is able to identify and align similar shapes and sub-shapes' is thus supported by in-sample successes and a maximized internal metric.

  2. self definitional [Results, 'Optimization of parameters for evaluation'; Methods III, Eq. 8]
    ""Since the distance between two points in our point clouds is 0.3, we set a threshold value to 0.3. Increasing this threshold would automatically and artificially increase gfit scores by pairing more distant points." "gfit = number-of-matching-points-in-Source / Total-number-of-points-in-Source""

    The quantity offered as evidence of alignment quality is the same quantity optimized by the voxel-grid loop, and its threshold is chosen by the authors from the point-cloud spacing. Without a null distribution of gfit for random placements, a high gfit can mean containment (small fragment inside a larger target), not a chemically correct match; the paper's own Adapalene/tetrazole alternate pose scores higher gfit (0.612 vs 0.603) and is only rejected by hfit. Hence gfit alone is a self-defined score, not external validation.

full rationale

The paper does not invoke any load-bearing self-citation or uniqueness theorem; the Open3D methods are external, machine-implemented algorithms. The circularity is limited to the evaluation layer: parameters (voxel grid, RANSAC iterations, COLOR iterations, evaluation threshold) were optimized on the same molecular sets that are later presented as successful validations, and the reported COLOR gfit is the best value selected from eleven down-sampling trials. This is in-sample tuning plus objective-maximization rather than a derivation, so it raises the circularity score but does not erase the independent content of the visual alignments and the ROCS comparison. Score 5 reflects partial, not total, circularity.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

No new physical entities are introduced. The central claim rests on a chain of domain assumptions: that vdW surface point clouds colored by four atom classes encode the relevant chemistry, that computer-vision descriptors work at molecular scale, and that the hand-tuned parameters generalize beyond the test cases used to set them. The GLOBAL, COLOR, and evaluation thresholds are all hand-chosen and directly affect the reported scores.

free parameters (5)
  • evaluation distance threshold = 0.3
    Set equal to the average spacing between surface points; the authors note that increasing it would artificially inflate gfit scores. This threshold directly defines what counts as a matched point.
  • voxel size grid = 0.2 to 1.2 in 0.1 increments
    Eleven down-sampling levels are tried because no single value worked for all test cases. The choice was made based on behavior on Imatinib and Sorbate moved-molecule pairs.
  • RANSAC max_iteration = 400000
    Reduced from the Open3D example value of 4,000,000 to keep computation time acceptable; chosen by hand.
  • RANSAC max_validation = 1000
    Increased from the Open3D default of 500; chosen by hand to improve validation success.
  • COLOR max_iteration = 100
    Values larger than 100 did not improve results in test cases; chosen by hand.
assumptions (5)
  • domain assumption The van der Waals surface computed by nsc with Bondi radii is an adequate representation of molecular shape for similarity comparison.
    Used throughout; no comparison to solvent-excluded surfaces or other surface definitions is provided.
  • domain assumption Coloring each surface point by the element class of its closest atom approximates the physicochemical features relevant to molecular recognition.
    Section I-3 asserts that the four classes recapitulate pharmacophore features, but this is not validated independently.
  • domain assumption FPFH descriptors and RANSAC matching, as implemented in Open3D, transfer correctly from indoor scenes to molecular-scale point clouds.
    The methods are imported without adaptation except for parameter changes; no theoretical justification for scale invariance is given.
  • standard math The Open3D library implementations of global registration, colored ICP, and evaluate registration are correct.
    Assumed from citation [9]; the authors do not verify the underlying algorithms.
  • ad hoc to paper The gfit, cfit, and hfit matching fractions are meaningful measures of molecular similarity.
    Defined in Section III; they are simple fractions of matched points under a fixed distance threshold and are not validated against bioactivity or external similarity judgments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SENSAAS (SENsitive Surface As A Shape): utilizing open-source algorithms for 3D point cloud alignment of molecules." pith.science (2026). https://pith.science/paper/QKJEU3XW

@misc{pith2026190811267,
  author       = {Pith},
  title        = {Pith review of: SENSAAS (SENsitive Surface As A Shape): utilizing open-source algorithms for 3D point cloud alignment of molecules},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QKJEU3XW}},
  note         = {Machine review of arXiv:1908.11267}
}
read the original abstract

Open-source 3D data processing libraries originally developed for computer vision and pattern recognition are used to align and compare molecular shapes and sub-shapes. Here, a shape is represented by a set of points distributed on the van der Waals surface of molecules. Each point is colored by its closest atom, which itself belongs to a user defined class. The strength of this representation is that it allows for comparisons of point clouds of different kind of chemical entities: small molecules, peptides, proteins or cavities (the negative image of the

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages

  1. [1]

    and Ekins, S

    Kortagere, S., Krasowski, M.D. and Ekins, S. (2009) The importance of discerning shape in molecular pharmacology. Trends in pharmacological sciences, 30, 138-147

    work page 2009

  2. [2]

    Nicholls, A., McGaughey, G.B., Sheridan, R.P., Good, A.C., Warren, G., Mathieu, M., Muchmore, S.W., Brown, S.P., Grant, J.A., Haigh, J.A. et al. (2010) Molecular shape and medicinal chemistry: a perspective. Journal of medicinal chemistry, 53, 3862-3886

    work page 2010

  3. [3]

    and Bajorath, J

    Maggiora, G., Vogt, M., Stumpfe, D. and Bajorath, J. (2014) Molecular similarity in medicinal chemistry. Journal of medicinal chemistry, 57, 3186-3204

    work page 2014

  4. [4]

    and Zhang, K.Y.J

    Kumar, A. and Zhang, K.Y.J. (2018) Advances in the Development of Shape Similarity Methods and Their Application in Drug Discovery. Frontiers in chemistry, 6, 315

    work page 2018

  5. [5]

    and Hege, H.C

    Baum, D. and Hege, H.C . (2006), Computational Life Sciences: Second International Symposium, CompLife 2006, Vol. 4216, pp. 183-193

    work page 2006

  6. [6]

    and Flynn, P

    Campbell, R. and Flynn, P. (2001) A Survey Of Free -Form Object Representation and Recognition Techniques. Computer Vision and Image Understanding, 81, 166-210

    work page 2001

  7. [7]

    and Levoy, M

    Rusinkiewicz, S. and Levoy, M. (2001) Efficient variants of the ICP algorithm. Proceedings Third International Conference on 3- D Digital Imaging and Modeling, 145- 152

    work page 2001

  8. [8]

    and Weyn, M

    Bellekens, B., Spruyt, V., Berkvens, R. and Weyn, M. (2014) A sur vey of rigid 3D pointcloud registration algorithms. Fourth International Conference on Ambient Computing, Applications, Services and Technologies, Proceedings, 8-13

    work page 2014

Show all 32 references
  1. [9]

    -Y., Park, J

    Zhou, Q. -Y., Park, J. and Koltun, V. (2018) Open3D: A Modern Library for 3D Data Processing

    2018

  2. [10]

    (2010) e -LEA3D: a computational -aided drug design web server

    Douguet, D. (2010) e -LEA3D: a computational -aided drug design web server. Nucleic acids research, 38, W615-621

    2010

  3. [11]

    and Scharf, M

    Eisenhaber, F., Lijnzaad, P., Argos, P., Sander, C. and Scharf, M. (1995) The Double Cubic Lattice Method: Efficient Approaches to Numerical Integration of Surface Area and Volume and to Dot Surface Contouring of Molecular Assemblies. J. Computat. Chem., 16, 273-284

    1995

  4. [12]

    (1964) van der Waals Volumes and Radii

    Bondi, A. (1964) van der Waals Volumes and Radii. J. Phys. Chem., 68, 441-451

    1964

  5. [13]

    and Schmidtke, P

    Shinada, N.K., de Brevern, A.G. and Schmidtke, P. (2019) Halogens in Protein- Ligand Binding Mechanism: A Structural Perspective. Journal of medicinal chemistry

    2019

  6. [14]

    and Beetz, M

    Rusu, R.B., Blodow, N. and Beetz, M. (2009) Fast Point Feature Histograms (FPFH) for 3D Registration. International Conference on Robotics and Automation (ICRA)

    2009

  7. [15]

    and Bolles, R.C

    Fischler, M.A. and Bolles, R.C. (1981) Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography. Comm. Of the ACM, 24, 381-395

    1981

  8. [16]

    and Koltun, V

    Choi, S., Zhou, Q.- Y. and Koltun, V. (2015) Robust Reconstruction of Indoor Scenes. IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    2015

  9. [17]

    and Koltun, V

    Park, J., Zhou, Q.- Y. and Koltun, V. (2017) Colored Point Cloud Registration Revisited. IEEE International Conference on Computer Vision (ICCV), 143-152

    2017

  10. [18]

    and Bajorath, J

    Stumpfe, D. and Bajorath, J. (2011) Similarity searching. WIREs Comput Mol Sci, 1, 260- 282. 27

    2011

  11. [19]

    Falck, J.R., Koduru, S.R., Mohapatra, S., Manne, R., Atcha, K.R., Atcha, R., Manthati, V.L., Capdevila, J.H., Christian, S., Imig, J.D. et al. (2014) 14,15-Epoxyeicosa-5,8,11- trienoic Acid (14,15 -EET) surrogates: carboxylate modifications. Journal of medicinal chemistry, 57,...

    2014

  12. [20]

    and Weaver, D.F

    Matta, C.F., Arabi, A.A. and Weaver, D.F. (2010) The bioisosteric similarity of the tetrazole and carboxylate anions: clues from the topologies of the electrostatic potential and of the electron density. European journal of medicinal chemistry, 45, 1868-1872

    2010

  13. [21]

    (2014) Bioisosteres and Scaffold Hopping in Medicinal Chemistry

    Brown, N. (2014) Bioisosteres and Scaffold Hopping in Medicinal Chemistry. Molecular informatics, 33, 458-462

    2014

  14. [22]

    and Pickup, B.T

    Grant, J.A., Gallardo, M.A. and Pickup, B.T. (1996) A fast method of molecular shape comparison: A simple application of a Gaussian description of molecular shape. Journal of Computational Chemistry, 17, 1653-1666

    1996

  15. [23]

    and Nicholls, A

    Hawkins, P.C., Skillman, A.G. and Nicholls, A. (2007) Comparison of shape -matching and docking as virtual screening tools. Journal of medicinal chemistry, 50, 74-82

    2007

  16. [24]

    ROCS 3.1.0 ; OMEGA 2.4.3

    OpeneEyeScientificSoftware. ROCS 3.1.0 ; OMEGA 2.4.3

  17. [25]

    and Sadowski, J

    Gasteiger, J., Rudolph, C. and Sadowski, J. (1990) Automatic Generation of 3D -Atomic Coordinates for Organic Molecules. Tetrahedron Comp. Method, 3, 537-547

    1990

  18. [26]

    and Stahl, M.T

    Hawkins, P.C., Skillman, A.G., Warren, G.L., Ellingson, B.A. and Stahl, M.T. (2010) Conformer generation with OMEGA: algorithm and validation using high quality structures from the Protein Databank and Cambridge Structural Database. Journal of chemical information and modeling...

    2010

  19. [27]

    (2018) Data Sets Representative of the Structures and Experimental Properties of FDA-Approved Drugs

    Douguet, D. (2018) Data Sets Representative of the Structures and Experimental Properties of FDA-Approved Drugs. ACS medicinal chemistry letters, 9, 204-209

    2018

  20. [28]

    (2005), Computational Life Sciences: F irst International Symposium, CompLife 2005, pp

    Baum, D. (2005), Computational Life Sciences: F irst International Symposium, CompLife 2005, pp. 198-207

    2005

  21. [29]

    (2007) A Point -Based Algorithm for Multiple 3D Surface Alignment of Drug - Sized Molecules

    Baum, D. (2007) A Point -Based Algorithm for Multiple 3D Surface Alignment of Drug - Sized Molecules

    2007

  22. [30]

    and Del Rio, A

    Sanders, M.P., Barbosa, A.J., Zarzycka, B., Nicolaes, G.A., Klomp, J.P., de Vlieg, J. and Del Rio, A. (2012) Comparative analysis of pharmacophore screening tools. Journal of chemical information and modeling, 52, 1607-1620

    2012

  23. [31]

    and Ritchie, D.W

    Venkatraman, V., Perez -Nueno, V.I., Mavridis, L. and Ritchie, D.W. (2010) Comprehensive comparison of ligand-based virtual screening tools against the DUD data set reveals limitations of current 3D methods. Journal of chemical information and modeling, 50, 2079-2093

    2010

  24. [32]

    and Johnson, M.S

    Vainio, M.J., Puranen, J.S. and Johnson, M.S. (2009) ShaEP: molecular overlay based on shape and electrostatic potential. Journal of chemical information and modeling, 49, 492- 502. Figure1 a) f) 4412points 6382points d) 6490points b) c) g) e) 6490points Aromaticgreenpatch Fig...

    2009

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.