The DOI matches its Crossref record; 54 of the 54 references it cites were found in a registry or an index; none of the 27 sources its claims cite is checked for support yet.
A topology-based network tree for the prediction of protein–protein binding affinity changes following mutation
Open accessAbstract
The ability to predict protein–protein interactions is crucial to our understanding of a wide range of biological activities and functions in the human body, and for guiding drug discovery. Despite considerable efforts to develop suitable computational methods, predicting protein–protein interaction binding affinity changes following mutation (ΔΔG) remains a severe challenge. Algebraic topology, a champion in recent worldwide competitions for protein–ligand binding affinity predictions, is a promising approach to simplifying the complexity of biological structures. Here we introduce element- and site-specific persistent homology (a new branch of algebraic topology) to simplify the structural complexity of protein–protein complexes and embed crucial biological information into topological invariants. We also propose a new deep learning algorithm called NetTree to take advantage of convolutional neural networks and gradient-boosting trees. A topology-based network tree is constructed by integrating the topological representation and NetTree for predicting protein–protein interaction ΔΔG. Tests on major benchmark datasets indicate that the proposed topology-based network tree is an important improvement over the current state of the art in predicting ΔΔG. Persistent homology provides an efficient approach to simplifying the complexity of protein structure. Wang et al. combine this approach with convolutional neural networks and gradient-boosting trees to improve predictions of protein–protein interactions.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper introduces TopNetTree, combining element- and site-specific persistent homology with CNN-assisted gradient-boosting trees to predict protein–protein binding affinity changes upon mutation (ΔΔG). It reports improved Pearson correlations over prior methods on AB-Bind and SKEMPI benchmarks, including RP 0.65 on AB-Bind S645 and RP 0.82 on SKEMPI 2.0 S4947.
Key findings
- On the AB-Bind S645 set, TopNetTree achieved an RP of 0.65, about 22% better than the previous best RP of 0.53 from mCSM-AB; excluding 27 non-binders increased RP to 0.68, and a blind test on homology structures gave RP 0.55.
- On the SKEMPI S1131 set, the method was reported as about 15% more accurate than the best existing method in tenfold cross-validation.
- On SKEMPI 2.0, TopNetTree achieved RP 0.82 and r.m.s.e. 1.11 kcal mol−1 on S4947; RP 0.79 and r.m.s.e. 1.13 kcal mol−1 on S4169 versus mCSM-PPI2’s RP 0.76 and r.m.s.e. 1.19 kcal mol−1; and RP 0.85 and r.m.s.e. 1.11 kcal mol−1 on S8338 versus mCSM-PPI2’s RP 0.82 and r.m.s.e. 1.18 kcal mol−1.
- Cross-validation analyses showed inter-protein average/median RP of 0.508/0.541 and intra-protein leave-one-out average/median RP of 0.170/0.215; mutation-region RP values were 0.60 (core), 0.66 (rim), 0.66 (support), 0.65 (interior) and 0.48 (surface).
Methods
- The model uses element- and site-specific persistent homology: atoms are classified into mutation-site, neighbourhood, antibody, antigen and element-type subsets, with modified distance matrices excluding same-set interactions; H0 barcodes come from Vietoris–Rips filtration, while H1 and H2 barcodes come from alpha complex filtration.
- H0 barcodes are vectorized by binning the filtration range (for example, [0, 12] Å into 0.5 Å bins) and counting persistence intervals, birth events and death events; H1 and H2 barcodes are summarized by statistics such as sum, maximum, minimum, mean and standard deviation of bar lengths, birth values and death values.
- A CNN extracts higher-level features from H0; these are combined with H1/H2 statistics and fed into a gradient-boosting tree (GBT) for final prediction. The paper evaluates on AB-Bind S645, SKEMPI S2317/S1131, and SKEMPI 2.0 S4947/S4169/S8338/S787 using RP and r.m.s.e., with tenfold, inter-protein, intra-protein and blind-test protocols.
Limitations
- Stated: the characterization of birth and death may not be stable against different discretizations, so only H0 barcodes from Vietoris–Rips filtration are used; datasets are relatively small and complex; non-binders are outliers that strongly affect accuracy; intra-protein cross-validation is limited by small per-complex training sets (average 27 samples); surface mutations have small data and small average energy disturbance; model prediction variance is generally lower than experimental variance.
- Evident from reported results: performance varies strongly across protein families, with negative RP for 1FFW (−0.043) and 1YY9 (−0.068); inter-protein cross-validation (average RP 0.508) and blind tests (homology RP 0.55; S787 RP 0.53) are lower than the main tenfold results.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 of 27 cited sources checked · 0 supported · 0 flagged
- Findingown data
TopNetTree achieved an RP of 0.55 in a blind test on homology structures using the rest of the samples as the training set.
“We also applied a blind test on homology structures using the rest of the samples as the training set, achieving an RP of 0.55, as shown in Fig. 4.”
Quote in the paperSubstantial claim - Findingown data
The proposed topology-based network tree (TopNetTree) is an important improvement over the current state of the art in predicting protein-protein interaction ΔΔG on major benchmark datasets.
“Tests on major benchmark datasets indicate that the proposed topology-based network tree is an important improvement over the current state of the art in predicting ΔΔG.”
Quote in the paperSubstantial claim- The authors state this as an assumption: The individual learners in an ensemble are likely to make different mistakes, so summing weak learners reduces overall error.
- Rests on an assumption: PPI ΔΔG prediction data are highly complex due to 3D structures, whereas the datasets are relatively small.
- Findingown data
TopNetTree is about 22% more accurate than the best-existing score of RP = 0.53 (given by mCSM-AB) on the AB-Bind S645 dataset.
“Our method is about 22% more accurate than the best-existing score of RP = 0.53 (given by mCSM-AB), indicating the power of our TopNetTree.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Pires et al.21 Pires, D. E. & Ascher, D. B. mCSM-AB: a web server for predicting antibody–antigen affinity changes upon mutation with graph-based signatures. Nucleic Acids Res. 44, W469–W473 (2016). source
- Findingown data
TopNetTree is over 100% more accurate in terms of RP than non-machine learning methods such as Rosetta and bASA on the AB-Bind S645 dataset.
“In comparison with non-machine learning methods such as Rosetta and bASA, our method is over 100% more accurate in terms of RP indicating that our topology-based machine learning methods have a better predictive power for PPI systems.”
Quote in the paperSubstantial claim - Findingown data
Excluding non-binders from the AB-Bind S645 dataset increased TopNetTree's RP from 0.65 to 0.68.
“In our case, the RP increased from 0.65 to 0.68 for the same treatment as shown in Fig. 4.”
Quote in the paperNot a substantial claim
- Findingown data
TopNetTree achieved an RP of 0.65 on the AB-Bind S645 dataset.
“Our model achieved an RP of 0.65 on the AB-Bind S645 dataset, which is significantly better than those of other existing methods as shown in Table 1.”
Quote in the paperSubstantial claim - Findingown data
TopNetTree is about 15% more accurate than the best-existing method on the SKEMPI S1131 set.
“It is found that the proposed TopNetTree is about 15% more accurate than the best-existing method.”
Quote in the paperSubstantial claim - Findingown data
The model's prediction performance varies among different protein families.
“however, the performance of the model varies among different protein families”
Quote in the paperNot a substantial claim - Findingown data
Most mutations from polar to hydrophobic residues have a positive free energy change (for example, S to M).
“we find that most mutations from polar to hydrophobic residues have a positive free energy change (for example, S to M)”
Quote in the paperNot a substantial claim - Findingnone given
The ΔΔG of the 27 non-binders (−8 kcal mol−1) did not follow the distribution of the whole dataset.
“The ΔΔG of 27 non-binders (−8 kcal mol−1) did not follow the distribution of the whole dataset.”
Quote in the paperSubstantial claim - Findingown data
The model achieved average RP values of 0.60, 0.66, 0.66, 0.65 and 0.48 for the core, rim, support, interior and surface mutation regions, respectively.
“Average RP values of 0.60, 0.66, 0.66, 0.65 and 0.48 were achieved for the core, rim, support, interior and surface regions, respectively.”
Quote in the paperSubstantial claim - Findingown data
Surface mutations have an average energy change of less than 0.2 kcal mol−1 in the experimental data.
“On the other hand, the surface mutations have an average energy change of less than 0.2 kcal mol−1.”
Quote in the paperNot a substantial claim - Findingnone given
For the benchmark PPI dataset SKEMPI, the present method significantly outperforms the state-of-the-art in the literature.
“For another benchmark PPI dataset, SKEMPI, the present method significantly outperforms the state-of-the-art in the literature.”
Quote in the paperNot a substantial claim - Findingown data
The variance of the model predictions is generally lower than that of the experimental data.
“the variance of the model predictions is generally lower than the experimental data as shown in Fig. 5”
Quote in the paperNot a substantial claim - Findingreasoning
A mutation from charged residues to uncharged polar residues could lead to a negative energy change; for example, lysine to serine (K to S).
“We also observed that a mutation from charged residues to uncharged polar residues could lead to a negative energy change; for example, lysine to serine (K to S)”
Quote in the paperNot a substantial claim - Premisecitation
Xiong et al. selected a subset of 1,131 non-redundant interface single-point mutations (S1131) from SKEMPI S2317.
“Xiong et al. recently selected a subset of 1,131 non-redundant interface single-point mutations (denoted set S1131) from SKEMPI set S231740.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [40] Xiong, P., Zhang, C., Zheng, W. & Zhang, Y. Bindprofx: assessing mutation-induced binding affinity change by protein interface profiles with pseudo-counts. J. Mol. Biol. 429, 426–434 (2017). source
- Premisecitation
mCSM-AB achieved a Pearson correlation coefficient of 0.53 with tenfold cross-validation on a set of 645 single-point mutations.
“This method (named mCSM-AB) was shown to outperform the aforementioned physical methods yet only achieve a Pearson’s correlation coefficient (RP) of 0.53 with tenfold cross-validation on a set of 645 single-point mutations.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Pires et al.21 Pires, D. E. & Ascher, D. B. mCSM-AB: a web server for predicting antibody–antigen affinity changes upon mutation with graph-based signatures. Nucleic Acids Res. 44, W469–W473 (2016). source
- Interpretationown data
The model needs a diversity of training samples to achieve stable and consistent prediction quality.
“This result also implies that our model needs a diversity of training samples to achieve stable and consistent prediction quality”
Quote in the paperSubstantial claim - Interpretationown data
The model's performance is consistent among different mutation regions except for the surface region.
“This result shows that the performance is consistent among different mutation regions except for the surface region.”
Quote in the paperNot a substantial claim - Interpretationown data
The inter-protein-level cross-validation performance is comparable with the result of the blind test on homology models.
“This performance is comparable with the result of blind test on homology models (see Fig. 4)”
Quote in the paperNot a substantial claim - Interpretationown data
A possible reason for the lower intra-protein-level cross-validation performance is that the training set for each complex is too small, with only an average of 27 samples per complex.
“One possible reason for this behaviour is that the training set for each complex is too small with only an average of 27 samples per complex.”
Quote in the paperSubstantial claim - Study design and conductown data
A blind test on homology structures was applied using the rest of the samples as the training set.
“We also applied a blind test on homology structures using the rest of the samples as the training set, achieving an RP of 0.55, as shown in Fig. 4.”
Quote in the paperSubstantial claim - Study design and conductown data
A blind prediction test was performed on 787 mutations from 24 protein complexes shared between AB-Bind and SKEMPI 2.0, using S4947 excluding S787 as training.
“We further validated our method by the blind prediction of another subset of the AB-Bind database. As SKEMPI 2.0 contains entries in the AB-Bind dataset, we chose 24 protein complexes that appear in both AB-Bind and SKEMPI 2.0 datasets as the test set for 787 mutations (denoted as the S787 set). The S4947 set, excluding the S787 set, was used as the training set.”
Quote in the paperSubstantial claim - Premisenone given
Element-specific persistent homology can retain critical biological information during topological abstraction.
“Element-specific persistent homology can retain critical biological information during the topological abstraction.”
Quote in the paperNot a substantial claim - Premisenone given
Techniques designed for protein-ligand binding analysis could not be directly applied to PPIs due to biological differences and the different characteristics of available datasets.
“however, the techniques designed for protein–ligand binding analysis could not be directly applied to PPIs due to biological differences and the different characteristics of available datasets.”
Quote in the paperNot a substantial claim - Interpretationown data
Mutations from polar residues to hydrophobic residues would make the whole PPI system more stable.
“which means mutations from polar residues to hydrophobic residues would make the whole PPI system more stable”
Quote in the paperSubstantial claim - Premisecitation
The RP value between predicted ΔΔG values and experimental data in cross-validations of the commonly used PPI database AB-Bind is only 0.53.
“the RP value between predicted ΔΔG values and experimental data in cross-validations of a commonly used PPI database, AB-Bind10, is only 0.53.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [10] Sirin, S., Apgar, J. R., Bennett, E. M. & Keating, A. E. AB-bind: antibody binding mutational database for computational affinity predictions. Protein Sci. 25, 393–409 (2016). source
- Interpretationreasoning
Such mutations from charged residues to uncharged polar residues might have broken some charge-charge interaction pairs.
“which means such mutations might have broken some charge–charge interaction pairs”
Quote in the paperNot a substantial claim
- It may state two separate claims as one. 15
- The quote only partly states this claim. 3
- Probably not a substantial claim. 17
It may state two separate claims as one. The quote only partly states this claim.
Findingown dataTopNetTree achieved an average RP of 0.53 and r.m.s.e. of 1.45 kcal mol−1 on a blind test of 787 mutations from 24 protein complexes shared between AB-Bind and SKEMPI 2.0 (S787), using S4947 excluding S787 as training.
“We achieved an average RP of 0.53 and r.m.s.e. of 1.45 kcal mol−1 on this blind test (further details are given in the Supplementary Information).”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingown dataTopNetTree achieved an average RP of 0.82 and an r.m.s.e. of 1.11 kcal mol−1 for tenfold cross-validation on SKEMPI 2.0 set S4947.
“We tested set S4947 with the regular tenfold cross-validation, achieving an average RP of 0.82 and an r.m.s.e. of 1.11 kcal mol−1 for the tenfold cross-validation.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingown dataOn SKEMPI 2.0 set S4169, TopNetTree achieved an average RP of 0.79 and r.m.s.e. of 1.13 kcal mol−1, compared with mCSM-PPI2's average RP of 0.76 and r.m.s.e. of 1.19 kcal mol−1.
“For set S4169, we obtained an average RP of 0.79 and r.m.s.e. of 1.13 kcal mol−1, compared with the average RP of 0.76 and r.m.s.e. of 1.19 kcal mol−1 achieved by mCSM-PPI2.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites David et al.48
It may state two separate claims as one.
Findingown dataThe AB-Bind S645 set consists of about 20% stabilizing mutations and 80% destabilizing ones, and there are 27 non-binders in the whole dataset.
“This dataset, called the AB-Bind S645 set, consists of about 20% stabilizing mutations and 80% destabilizing ones; there are 27 non-binders in the whole dataset, which are variants determined not to bind within the sensitivity of the assay.”
Quote in the paperNot a substantial claimProbably not a substantial claim. The paper’s conclusions may not depend on it.
FindingreasoningAll mutations to alanine have a positive energy change.
“It is interesting to note that all the mutations to alanine have a positive energy change”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
FindingreasoningIn the experimental data, mutations at the core or support region have a higher average energy change of around 1.8 kcal mol−1 (1.72 and 1.91 kcal mol−1, respectively), whereas mutations at the rim or interior region have an average energy change of around 0.8 kcal mol−1 (0.82 and 0.83 kcal mol−1, respectively).
“In experimental data, mutations at the core or support region have a higher average energy change of around 1.8 kcal mol−1 (1.72 kcal mol−1 and 1.91 kcal mol−1, respectively), whereas mutations at the rim or interior region have an average energy change of around 0.8 kcal mol−1 (0.82 kcal mol−1 and 0.83 kcal mol−1, respectively)”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataThe model's predicted patterns are remarkably similar to those of experimental data in terms of both average binding energy changes and variance of binding energy changes.
“Overall, our predicted patterns are remarkably similar to those of experimental data in terms of both average binding energy changes and variance of binding energy changes, as shown in Fig. 5.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
FindingreasoningIn intra-protein-level leave-one-out cross-validation within each protein complex, the model reached average/median RP values of 0.170/0.215, significantly lower than the tenfold cross-validation result over the entire dataset.
“For this test, our model reached average/median RP values of 0.170/0.215, which are significantly lower than the tenfold cross-validation result over the entire dataset.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataIn inter-protein-level cross-validation over the 24 protein-protein complexes of the AB-Bind S645 set, the model reached an average RP of 0.508 and a median RP of 0.541.
“For this test, our model reached an average RP of 0.508 and a median RP of 0.541.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingown dataThe proposed TopNetTree achieves an RP of 0.65, which is about 22% better than the previous best result for the AB-Bind dataset.
“We demonstrate that the proposed TopNetTree achieves an RP of 0.65, which is about 22% better than the previous best result for the AB-Bind dataset.”
Quote in the paperSubstantial claimProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
PremisecitationThe AB-Bind database contains 1,101 mutation data entries, including 645 single-point mutations on 32 different antibody-antigen complexes.
“Sirin et al.10 collected an AB-Bind database of mutation-induced antibody–antigen complex binding free energy changes. This database contains 1,101 mutation data entries, including 645 single-point mutations on 32 different antibody–antigen complexes.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Sirin et al.10 Sirin, S., Apgar, J. R., Bennett, E. M. & Keating, A. E. AB-bind: antibody binding mutational database for computational affinity predictions. Protein Sci. 25, 393–409 (2016). source
Probably not a substantial claim.
PremisecitationThe SKEMPI dataset contains 2,317 single point mutation data entries (SKEMPI S2317 set).
“There are 2,317 single point mutation data entries among the whole database, which are referred to as the SKEMPI S2317 set.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [11] Moal, I. H. & Fernández-Recio, J. SKEMPI: a structural kinetic and energetic database of mutant protein interactions and its use in empirical models. Bioinformatics 28, 2600–2607 (2012). source
Probably not a substantial claim.
PremisecitationSKEMPI 2.0 contains 7,085 entries, including single- and multi-point mutations.
“This dataset contains 7,085 entries, including single- and multi-point mutations.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [39] Jankauskaitė, J., Jiménez-García, B., Dapkūnas, J., Fernández-Recio, J. & Moal, I. H. SKEMPI 2.0: an updated benchmark of changes in protein–protein binding energy, kinetics and thermodynamics upon mutation. Bioinformatics 35, 462–469 (2018). source
Probably not a substantial claim. The paper’s conclusions may not depend on it.
PremisecitationPires et al. found that excluding non-binders from the dataset would significantly increase the performance of a prediction model.
“Pires et al.21 found that excluding non-binders from the dataset would significantly increase the performance of a prediction model.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Pires et al.21 Pires, D. E. & Ascher, D. B. mCSM-AB: a web server for predicting antibody–antigen affinity changes upon mutation with graph-based signatures. Nucleic Acids Res. 44, W469–W473 (2016). source
Probably not a substantial claim. The paper’s conclusions may not depend on it.
PremisecitationSKEMPI contains 3,047 mutation data entries for protein-protein heterodimeric complexes with experimentally determined structures.
“SKEMPI is a more general database for protein–protein binding affinity changes following mutation (ΔΔG)11, it contains 3,047 mutation data entries for protein–protein heterodimeric complexes with experimentally determined structures.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [11] Moal, I. H. & Fernández-Recio, J. SKEMPI: a structural kinetic and energetic database of mutant protein interactions and its use in empirical models. Bioinformatics 28, 2600–2607 (2012). source
Probably not a substantial claim. It may be too vague to check against evidence.
InterpretationreasoningThe limited performance of current methods highlights a pressing need for a new generation of ΔΔG predictors constructed with entirely new design principles and/or innovative machine learning algorithms.
“The limited performance of the current methods therefore highlights a pressing need for a new generation of ΔΔG predictors that are constructed with entirely new design principles and/or innovative machine learning algorithms.”
Quote in the paperNot a substantial claimProbably not a substantial claim. The paper’s conclusions may not depend on it.
InterpretationreasoningA possible reason for the positive energy change of mutations to alanine is that mutations from a large residue to a small one could lead to a stabilizing effect on the whole system.
“a possible reason is that mutations from a large residue to a small one could lead to a stabilizing effect to the whole system”
Quote in the paperNot a substantial claimProbably not a substantial claim. It may state two separate claims as one.
PremisecitationPersistent homology is able to bridge geometry and topology, leading to a new efficient approach for the simplification of biological structural complexity.
“Persistent homology22–25—a new branch of algebraic topology—is able to bridge geometry and topology, leading to a new efficient approach for the simplification of biological structural complexity26–31; however, it neglects critical chemical/biological information when it is directly applied to complex biomolecular structures.”
Quote in the paperSubstantial claim0 of 10 sources checked- cites [22] Frosini, P. & Landi, C. Size theory as a topological tool for computer vision. Pattern Recognition Image Anal. 9, 596–603 (1999).
- cites [23] Edelsbrunner, H., Letscher, D. & Zomorodian, A. Topological persistence and simplification. Discrete Comput. Geom. 28, 511–533 (2002). source
- cites [24] Zomorodian, A. & Carlsson, G. Computing persistent homology. Discrete Comput. Geom. 33, 249–274 (2005). source
- cites [25] Zomorodian, A. & Carlsson, G. Localized homology. Comput. Geom. 41, 126–148 (2008). source
- cites [26] Xia, K. L. & Wei, G. W. Persistent homology analysis of protein structure, flexibility and folding. Int. J. Numer. Methods Biomed. Eng. 30, 814–844 (2014). source
- cites [27] Gameiro, M. et al. Topological measurement of protein compressibility via persistence diagrams. Japan J. Industr. Appl. Math. 32, 1–17 (2014).
- cites [28] Xia, K. L. & Wei, G. W. Persistent topology for cryo-EM data analysis. Int. J. Numer. Methods Biomed. Eng. 31, e02719 (2015).
- cites [29] Cang, Z. X. et al. A topological approach to protein classification. Mol. Based Math. Biol. 3, 140–162 (2015).
- cites [30] Yao, Y. et al. Topological methods for exploring low-density states in biomolecular folding pathways. J. Chem. Phys. 130, 04B614 (2009).
- cites [31] Kovacev-Nikolic, V., Bubenik, P., Nikolić, D. & Heo, G. Using persistent homology and dynamical distances to analyze protein binding. Stat. Appl. Genet. Mol. Biol. 15, 19–38 (2016). source
It may state two separate claims as one.
Interpretationnone givenThe non-binders in the AB-Bind dataset can be regarded as outliers and have a strongly negative impact on prediction model accuracy.
“These non-binders could be regarded as outliers in the database and have a strongly negative impact on the prediction model accuracy.”
Quote in the paperNot a substantial claimProbably not a substantial claim. The paper’s conclusions may not depend on it.
InterpretationreasoningA possible reason for the observed patterns of energy change across mutation regions is that different mutation regions vary in their accessibility to water.
“A possible reason for these patterns is that different mutation regions vary in their accessibility to water”
Quote in the paperNot a substantial claim- Rests on an assumption: In general, surface, interior and rim regions have greater access to water than the core and support regions.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
PremisecitationSimilar patterns regarding mutation sites and energy changes were reported in the literature.
“Similar patterns regarding mutation sites and energy changes were reported in the literature49.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [49] Petukh, M., Li, M. & Alexov, E. Predicting binding free energy change caused by point mutations with knowledge-modified MM/PBSA method. PLoS Comput. Biol. 11, e1004276 (2015). source
Probably not a substantial claim. The quote only partly states this claim.
Study design and conductcitationFor sets S4169 and S8338, tenfold stratified cross-validation was used following the mCSM-PPI2 paper.
“We followed the method of tenfold stratified cross-validation used in mCSM-PPI2 paper for sets S4169 and S833848.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [48] Rodrigues, C. H. M., Myung, Y., Pires, D. E. V. & Ascher, D. B. mCSM-PPI2: Predicting the effects of mutations on protein–protein interactions. Nucleic Acids Res. 47, W338– W344 (2019). source
Probably not a substantial claim. The quote only partly states this claim.
Study design and conductown dataThe performance on the SKEMPI S1131 set was assessed using tenfold cross-validation.
“Table 2 shows the Pearson correlation coefficients on tenfold cross-validations.”
Quote in the paperSubstantial claimProbably not a substantial claim.
PremisecitationDavid et al. selected 4,169 variants in 319 different complexes (S4169) from SKEMPI2 by filtering only single-point mutations.
“They filtered only single-point mutations and selected 4,169 variants in 319 different complexes (denoted set S4169).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [48] Rodrigues, C. H. M., Myung, Y., Pires, D. E. V. & Ascher, D. B. mCSM-PPI2: Predicting the effects of mutations on protein–protein interactions. Nucleic Acids Res. 47, W338– W344 (2019). source
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
Premisenone givenPhysics-based methods provide mechanistic interpretations but are not designed for handling large and diverse datasets.
“Typically, physics-based methods provide mechanistic interpretations but are not designed for handling large and diverse datasets.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Interpretationown dataThe relative inferior performance for surface mutations is due to its small data size and the small average energy disturbance caused by surface mutations.
“We believe that the relative inferior performance for surface mutations is due to its small data size and that the energy disturbance caused by surface mutations is small on average.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Interpretationown dataModels trained on some protein families could extrapolate to other families, with the best families 1KTZ and 2JEL reaching RP of 0.866 and 0.818 and the poorest families 1FFW and 1YY9 having RP of −0.043 and −0.068.
“Models trained on some protein families could extrapolate to other families; for example, the two protein families with the best results, 1KTZ and 2JEL, can reach RP of 0.866 and 0.818, respectively, whereas the two families with the poorest results, 1FFW and 1YY9, have RP values of −0.043 and −0.068, respectively.”
Quote in the paperNot a substantial claimProbably not a substantial claim.
PremisecitationTopology has recently been shown to be surprisingly effective in simplifying biomolecular structural complexity.
“Topology has recently been shown to be surprisingly effective in simplifying biomolecular structural complexity26,27,29.”
Quote in the paperSubstantial claim0 of 3 sources checked- cites [26] Xia, K. L. & Wei, G. W. Persistent homology analysis of protein structure, flexibility and folding. Int. J. Numer. Methods Biomed. Eng. 30, 814–844 (2014). source
- cites [27] Gameiro, M. et al. Topological measurement of protein compressibility via persistence diagrams. Japan J. Industr. Appl. Math. 32, 1–17 (2014).
- cites [29] Cang, Z. X. et al. A topological approach to protein classification. Mol. Based Math. Biol. 3, 140–162 (2015).
Probably not a substantial claim. The paper’s conclusions may not depend on it.
PremisecitationTopology has been devised to win worldwide competitions in computer-aided drug design.
“It has been devised to win worldwide competitions in computer-aided drug design36.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [36] Nguyen, D. D. et al. Mathematical deep learning for pose and binding affinity prediction and ranking in D3R grand challenges. J. Compurt. Aided Mol. Design https://doi.org/10.1007/s10822-018-0146-6 (2018). source
The authors state this as an assumption.
The individual learners in an ensemble are likely to make different mistakes, so summing weak learners reduces overall error.
“It improves the performance on the weak learners with the assumption that the individual learners are likely to make different mistakes and thus summing the weak learners will reduce the overall error.”
The paper relies on this without support.
PPI ΔΔG prediction data are highly complex due to 3D structures, whereas the datasets are relatively small.
“A major challenge in the prediction of binding affinity changes following mutation for PPIs is that the data is highly complex due to 3D structures, whereas the datasets are relatively small.”
The paper relies on this without support.
In general, surface, interior and rim regions have greater access to water than the core and support regions.
“in general, surface, interior and rim regions have greater access to water than the core and support regions.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
TopGBT uses topology-based features not initially designed for representing interactions between atoms, limiting its predictive power for binding affinity changes upon mutations.
“TopGBT uses topology-based features to represent the complex, which is not initially designed for representing the interactions between atoms, limiting its predictive power for binding affinity changes upon mutations.”
From Deep geometric representations for modeling effects of mutations on protein-protein binding affinity (2021) · cites this paper as [31]
Some complexes in the benchmark datasets are highly related, so machine learning methods tend to be overtrained on these datasets, as evidenced by drops in prediction performance when meeting unseen structures.
“As some complexes in the above benchmark datasets are highly related (S2 and S3 Figs), machine learning methods tend to be overtrained on these datasets, as evidenced by the drops of the prediction performance when machine learning methods meet unseen structures [31].”
From Deep geometric representations for modeling effects of mutations on protein-protein binding affinity (2021) · cites this paper as [31]
The six benchmark datasets have been widely used to test the predictive power of PPI prediction methods.
“These datasets have been widely used to test the predictive power of the PPI prediction methods [12, 13, 27–29, 29–35].”
From Deep geometric representations for modeling effects of mutations on protein-protein binding affinity (2021) · cites this paper as [31]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
The same 645 AB-Bind mutations are assigned to different numbers of complexes
MinorThe paper first says the 645 single-point mutations are on 32 antibody-antigen complexes, then says it considers 645 single mutations across 29 complexes. It does not explain a change in the number of mutations that would account for the different complex counts.
“This database contains 1,101 mutation data entries, including 645 single-point mutations on 32 different antibody–antigen complexes.”
“We follow Pires et al.21 by considering only 645 single mutations across 29 antibody–antigen complexes.”
Citations
Cites
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016 · cited 1×
“The implementation provided by the scikit-learn package (v.0.18.1) 38 is used.”
MethodInfluentialBerman, H. M. (2000) · Nucleic Acids Research · doi:10.1093/nar/28.1.235 · Cited by 40,196
The CCP4 suite: programs for protein crystallography
Number Collaborative Computational Project (1994) · Acta Crystallographica Section D Biological Crystallography · doi:10.1107/s0907444994003112 · Cited by 17,317
Comparative Protein Structure Modeling Using MODELLER
Webb, B., & Šali, A. (2016) · Current Protocols in Bioinformatics · doi:10.1002/cpbi.3 · Cited by 4,404 · cited 1×
“Computational methods for generating protein structures from sequences (for example, MODELLER 16 ) and predicting docking poses for protein–protein complexes (for example, BioLuminate 17 ) are also available.”
MethodComparative Protein Structure Modeling Using MODELLER
Eswar, N., Webb, B., Martı́-Renom, M. A., Madhusudhan, M. S., Eramian, D., Shen, M., Pieper, U., & Šali, A. (2007) · Current Protocols in Protein Science · doi:10.1002/0471140864.ps0209s50 · Cited by 2,874
The FoldX web server: an online force field
Schymkowitz, J., Borg, J., Stricher, F., Nys, R., Rousseau, F., & Serrano, L. (2005) · Nucleic Acids Research · doi:10.1093/nar/gki387 · Cited by 2,759
Topological Persistence and Simplification
Edelsbrunner, Letscher, & Zomorodian (2002) · Discrete & Computational Geometry · doi:10.1007/s00454-002-2885-2 · Cited by 1,779
Zomorodian, A., & Carlsson, G. (2005) · Discrete & Computational Geometry · doi:10.1007/s00454-004-1146-y · Cited by 1,591
Conformations of immunoglobulin hypervariable regions
Chothia, C., Lesk, A. M., Tramontano, A. L., Levitt, M., Smith‐Gill, S. J., Air, G. M., Sheriff, S., PADLAN, E. A., Davies, D. R., Tulip, W. R., et al. (1989) · Nature · doi:10.1038/342877a0 · Cited by 1,191
Comparative Protein Structure Modeling Using MODELLER
Webb, B., & Šali, A. (2014) · Current Protocols in Bioinformatics · doi:10.1002/0471250953.bi0506s47 · Cited by 1,156
Potent antibody therapeutics by design
Carter, P. J. (2006) · Nature reviews. Immunology · doi:10.1038/nri1837 · Cited by 1,124
Interior and surface of monomeric proteins
Miller, S., Janin, J., Lesk, A. M., & Chothia, C. (1987) · Journal of Molecular Biology · doi:10.1016/0022-2836(87)90038-6 · Cited by 952
Challenges in the development of high protein concentration formulations
Shire, S. J., Shahrokh, Z., & 刘, J. 军. L. (2004) · Journal of Pharmaceutical Sciences · doi:10.1002/jps.20079 · Cited by 886
Challenges in the Development of High Protein Concentration Formulations
Shire, S. J., Shahrokh, Z., & Liu, J. (2009) · Springer eBooks · doi:10.1007/978-0-387-76643-0_9 · Cited by 657
Barouch, D. H., Whitney, J. B., Moldt, B., Klein, F., Oliveira, T. Y., Liu, J., Stephenson, K. E., Chang, H., Shekhar, K., Gupta, S., et al. (2013) · Nature · doi:10.1038/nature12744 · Cited by 644
Kortemme, T., Morozov, A. V., & Baker, D. A. (2003) · Journal of Molecular Biology · doi:10.1016/s0022-2836(03)00021-4 · Cited by 516
Jankauskaitė, J., Jiménez-García, B., Dapkūnas, J., Fernández-Recio, J., & Moal, I. H. (2019) · Bioinformatics · doi:10.1093/bioinformatics/bty635 · Cited by 389
mCSM-PPI2: predicting the effects of mutations on protein–protein interactions
Rodrigues, C. H. M., Myung, Y., Pires, D. E. V., & Ascher, D. B. (2019) · Nucleic Acids Research · doi:10.1093/nar/gkz383 · Cited by 369
BeAtMuSiC: prediction of changes in protein–protein binding affinity on mutations
Dehouck, Y., Kwasigroch, J. M., Rooman, M., & Gilis, D. (2013) · Nucleic Acids Research · doi:10.1093/nar/gkt450 · Cited by 338
Cang, Z., & Wei, G.-W. (2017) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1005690 · Cited by 330
Self-citationMoal, I. H., & Fernández-Recio, J. (2012) · Bioinformatics · doi:10.1093/bioinformatics/bts489 · Cited by 281
Cang, Z., Mu, L., & Wei, G.-W. (2018) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1005929 · Cited by 257
Self-citationA simple definition of structural regions in proteins and its use in analyzing interface evolution
Levy, E. D. (2010) · Journal of Molecular Biology · doi:10.1016/j.jmb.2010.09.028 · Cited by 254
Persistent homology analysis of protein structure, flexibility, and folding
Xia, K., & Wei, G. (2014) · International Journal for Numerical Methods in Biomedical Engineering · doi:10.1002/cnm.2655 · Cited by 254
Self-citationDocking, scoring, and affinity prediction in CAPRI
Lensink, M. F., & Wodak, S. J. (2013) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.24428 · Cited by 249
Zhu, K., Day, T., Warshaviak, D., Murrett, C. S., Friesner, R., & Pearlman, D. (2014) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.24551 · Cited by 206
Liu, S., Zhang, C., Zhou, H., & Zhou, Y. (2004) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.20019 · Cited by 193
AB ‐Bind: Antibody binding mutational database for computational affinity predictions
Sirin, S., Apgar, J. R., Bennett, E. M., & Keating, A. E. (2016) · Protein Science · doi:10.1002/pro.2829 · Cited by 179
Cang, Z., & Wei, G. (2018) · International Journal for Numerical Methods in Biomedical Engineering · doi:10.1002/cnm.2914 · Cited by 172
Self-citationTemplate-based structure modeling of protein–protein interactions
Szilágyi, A., & Zhang, Y. (2013) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2013.11.005 · Cited by 168
Renaissance of cancer therapeutic antibodies
Glennie, M. J., & van de Winkel, J. G. J. (2003) · Drug Discovery Today · doi:10.1016/s1359-6446(03)02714-4 · Cited by 155
Nguyen, D. D., Cang, Z., Wu, K., Wang, M., Cao, Y., & Wei, G.-W. (2019) · Journal of Computer-Aided Molecular Design · doi:10.1007/s10822-018-0146-6 · Cited by 155
Self-citationXiong, P., Zhang, C., Zheng, W., & Zhang, Y. (2017) · Journal of Molecular Biology · doi:10.1016/j.jmb.2016.11.022 · Cited by 140
Size theory as a topological tool for computer vision
Frosini, P., & Landi, C. (1999) · Cited by 137
A topological measurement of protein compressibility
Gameiro, M., Hiraoka, Y., Izumi, S., Kramár, M., Mischaikow, K., & Nanda, V. (2014) · Japan Journal of Industrial and Applied Mathematics · doi:10.1007/s13160-014-0153-5 · Cited by 134
Pires, D. E. V., & Ascher, D. B. (2016) · Nucleic Acids Research · doi:10.1093/nar/gkw458 · Cited by 134
Antibody therapeutics, antibody engineering, and the merits of protein stability.
Demarest, S. J., & Glaser, S. (2008) · PubMed · Cited by 131
Topological methods for exploring low-density states in biomolecular folding pathways
Yao, Y., Sun, J. Z., Huang, X., Bowman, G. R., Singh, G., Lesnick, M., Guibas, L. J., Pande, V. S., & Carlsson, G. (2009) · The Journal of Chemical Physics · doi:10.1063/1.3103496 · Cited by 129
Accurate and efficient loop selections by the DFIRE‐based all‐atom statistical potential
Zhang, C., Liu, S., & Zhou, Y. (2004) · Protein Science · doi:10.1110/ps.03411904 · Cited by 120
Using persistent homology and dynamical distances to analyze protein binding
Kovacev-Nikolic, V., Bubenik, P., Nikolić, D., & Heo, G. (2016) · Statistical Applications in Genetics and Molecular Biology · doi:10.1515/sagmb-2015-0057 · Cited by 116
Petukh, M., Li, M., & Alexov, E. (2015) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1004276 · Cited by 107
A distance for similarity classes of submanifolds of a Euclidean space
Frosini, P. (1990) · Bulletin of the Australian Mathematical Society · doi:10.1017/s0004972700028574 · Cited by 107
Geng, C., Xue, L. C., Roel‐Touris, J., & Bonvin, A. M. J. J. (2019) · Wiley Interdisciplinary Reviews Computational Molecular Science · doi:10.1002/wcms.1410 · Cited by 104
Cang, Z., & Wei, G. (2017) · Computer applications in the biosciences · doi:10.1093/bioinformatics/btx460 · Cited by 95 · cited 1×
“Paired with advanced machine learning, such as a convolutional neural network (CNN), this new topological method gives rise to some of the best predictions for protein–ligand binding affinities 32 , protein folding free energy changes following mutation 33,34 and drug virtual screening 35 .”
MethodSelf-citationA topological approach for protein classification
Cang, Z., Mu, L., Wu, K., Opron, K., Xia, K., & Wei, G.-W. (2015) · Computational and Mathematical Biophysics · doi:10.1515/mlbmb-2015-0009 · Cited by 93 · cited 1×
“Topology has recently been shown to be surprisingly effective in simplifying biomolecular structural complexity 26,27,29 .”
BackgroundSelf-citationZomorodian, A., & Carlsson, G. (2008) · Computational Geometry · doi:10.1016/j.comgeo.2008.02.003 · Cited by 89
PROXiMATE: a database of mutant protein–protein complex thermodynamics and kinetics
Jemimah, S., Yugandhar, K., & Gromiha, M. M. (2017) · Bioinformatics · doi:10.1093/bioinformatics/btx312 · Cited by 73
Structure-Based Predictive Models for Allosteric Hot Spots
Demerdash, O., Daily, M. D., & Mitchell, J. C. (2009) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1000531 · Cited by 67
SAAMBE: webserver to predict the charge of binding free energy caused by amino acids mutations
Petukh, M., Dai, L., & Alexov, E. (2016) · International Journal of Molecular Sciences · doi:10.3390/ijms17040547 · Cited by 65
Cancer therapeutic antibodies come of age: Targeting minimal residual disease
Ben-Kasus, T., Schechter, B., Sela, M., & Yarden, Y. (2007) · Molecular Oncology · doi:10.1016/j.molonc.2007.01.003 · Cited by 59
dbMPIKT: a database of kinetic and thermodynamic mutant protein interactions
Liu, Q., Chen, P., Wang, B., Zhang, J., & Li, J. (2018) · BMC Bioinformatics · doi:10.1186/s12859-018-2493-7 · Cited by 38
Persistent topology for cryo‐EM data analysis
Xia, K., & Wei, G. (2015) · International Journal for Numerical Methods in Biomedical Engineering · doi:10.1002/cnm.2719 · Cited by 36
Self-citationTopological Measurement of Protein Compressibility via Persistence DiagramsPreprint
Gameiro, M., ガメイロ, マ., Hiraoka, Y., 平岡, 裕., ヒラオカ, ヤ., Izumi, S., 泉, 俊., イズミ, シ., Kramár, M., クラマー, ミ., et al. (2012) · Kyushu University Institutional Repository (QIR) (Kyushu University) · Cited by 30
Prospective virtual screening for novel p53–MDM2 inhibitors using ultrafast shape recognition
Patil, S. P., Ballester, P. J., & Kerezsi, C. R. (2014) · Journal of Computer-Aided Molecular Design · doi:10.1007/s10822-014-9732-4 · Cited by 26
Flagged first, then by citations
Cited by
Sorted by citations
Omicron Variant (B.1.1.529): Infectivity, Vaccine Breakthrough, and Antibody Resistance
Chen, J., Wang, R., Gilby, N. B., & Wei, G. (2022) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.1c01451 · Cited by 709
Self-citationMutations Strengthened SARS-CoV-2 Infectivity
Chen, J., Wang, R., Wang, M., & Wei, G. (2020) · Journal of Molecular Biology · doi:10.1016/j.jmb.2020.07.009 · Cited by 468
“The resulting method, called TopNetTree, was about 22% better than the previous best result for the AB-Bind dataset and significantly outperformed the state-ofthe-art in the literature on the SKEMPI database [29].”
BackgroundMethodInfluentialSelf-citationDeep learning methods for molecular representation and property prediction
Li, Z., Jiang, M., Wang, S., & Zhang, S. (2022) · Drug Discovery Today · doi:10.1016/j.drudis.2022.103373 · Cited by 289
Wang, R., Chen, J., Gao, K., & Wei, G.-W. (2021) · Genomics · doi:10.1016/j.ygeno.2021.05.006 · Cited by 191
Self-citationWang, R., Chen, J., Gao, K., Hozumi, Y., Yin, C., & Wei, G. (2021) · Communications Biology · doi:10.1038/s42003-021-01754-6 · Cited by 201
Self-citationShan, S., Luo, S., Yang, Z., Hong, J., Su, Y., Ding, F., Fu, L., Li, C., Chen, P., Ma, J., et al. (2022) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2122954119 · Cited by 157
Xu, R., Cao, J., Feng, G., Luo, J., Feng, Q., Ni, B., & Fang, F. (2021) · Chemical Engineering Journal · doi:10.1016/j.cej.2021.132893 · Cited by 146
Deep geometric representations for modeling effects of mutations on protein-protein binding affinity
Liu, X., Luo, Y., Li, P., Song, S., & Peng, J. (2021) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1009284 · Cited by 140
“The methods that obtained the best results in literature on these two settings are TopGBT (Wang et al., 2020) and MutaBind2 (Zhang et al., 2020), respectively, which were used as the main base-lines in the comparison with GraphPPI.”
MethodAkbar, R., Bashour, H., Rawat, P., Robert, P. A., Smorodina, E., Cotet, T., Karlsen, K. F., Frank, R., Mehta, B. B., Vu, M. H., et al. (2022) · mAbs · doi:10.1080/19420862.2021.2008790 · Cited by 120
“Topology-based network tree (TopNetTree) was developed to predict changes in protein–protein interaction (PPI) affinity upon engineering, 174 and was built by combining the CNN with gradient-boosting trees (GBT).”
MethodPrediction and mitigation of mutation threats to COVID-19 vaccines and antibody therapies
Chen, J., Gao, K., Wang, R., & Wei, G. (2021) · Chemical Science · doi:10.1039/d1sc01203g · Cited by 110
Self-citationSAAMBE-3D: Predicting Effect of Mutations on Protein–Protein Interactions
Pahari, S., Li, G., Murthy, A. K., Liang, S., Fragoza, R., Yu, H., & Alexov, E. (2020) · International Journal of Molecular Sciences · doi:10.3390/ijms21072563 · Cited by 111
“TopNetTree [51] is another new method that utilized a topology-based network tree by integrating a deep learning algorithm, NetTree, and the topological representation.”
MethodAdvances in computational structure-based antibody design
Hummer, A. M., Abanades, B., & Deane, C. M. (2022) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2022.102379 · Cited by 109
Mechanisms of SARS-CoV-2 Evolution Revealing Vaccine-Resistant Mutations in Europe and America
Wang, R., Chen, J., & Wei, G. (2021) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.1c03380 · Cited by 110
Self-citationKolmogorov–Arnold graph neural networks for molecular property prediction
Li, L., Zhang, Y., Wang, G., & Xia, K. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-025-01087-7 · Cited by 130
“However, traditional covalent-bond-based molecular graph representations have various limitations, while incorporating non-covalent interactions has been shown to notably enhance performance 27 .”
Omicron BA.2 (B.1.1.529.2): High Potential for Becoming the Next Dominant Variant
Chen, J., & Wei, G. (2022) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.2c00469 · Cited by 103
Self-citationQiu, Y., & Wei, G. (2023) · Briefings in Bioinformatics · doi:10.1093/bib/bbad289 · Cited by 91
“There are several typical types of feature embedding methods, including sequence-based, structure-based [16, 17], physics-based [18, 19], and hybrid methods [20].”
Self-citationEngineering thermostability of industrial enzymes for enhanced application performance
Xu, K., Fu, H., Chen, Q., Sun, R., Li, R., Zhao, X., Zhou, J., & Wang, X. (2024) · International Journal of Biological Macromolecules · doi:10.1016/j.ijbiomac.2024.139067 · Cited by 88
Methodology-Centered Review of Molecular Modeling, Simulation, and Prediction of SARS-CoV-2
Gao, K., Wang, R., Chen, J., Cheng, L., Frishcosy, J., Huzumi, Y., Qiu, Y., Schluckbier, T., Wei, X., & Wei, G. (2022) · Chemical Reviews · doi:10.1021/acs.chemrev.1c00965 · Cited by 93
Self-citationChen, C., Boorla, V. S., Banerjee, D., Chowdhury, R., Cavener, V. S., Nissly, R. H., Gontu, A., Boyle, N. R., Vandegrift, K. J., Nair, M. S., et al. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2106480118 · Cited by 81
“However, the performance of the method used was tested on the general SKEMPI-2.”
MethodEmerging Vaccine-Breakthrough SARS-CoV-2 Variants
Wang, R., Chen, J., Hozumi, Y., Yin, C., & Wei, G. (2022) · ACS Infectious Diseases · doi:10.1021/acsinfecdis.1c00557 · Cited by 82
Self-citationWee, J., & Wei, G. (2024) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.4c00976 · Cited by 103
Self-citationOmicron (B.1.1.529): Infectivity, vaccine breakthrough, and antibody resistancePreprint
Chen, J., Wang, R., Gilby, N. B., & Wei, G. (2021) · PubMed · doi:10.48550/arxiv.2112.01318 · Cited by 78
“Mathematical features, including the element-specific and site-specific persistent homology (algebraic topology), are implemented to simplify the structural complexity of PPI complexes [9,28].”
BackgroundSelf-citationBig data: Historic advances and emerging trends in biomedical research
Cremin, C. J., Dash, S., & Huang, X. (2022) · Current Research in Biotechnology · doi:10.1016/j.crbiot.2022.02.004 · Cited by 91
Guarra, F., & Colombo, G. (2023) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.3c00513 · Cited by 82
A review of machine learning methods for imbalanced data challenges in chemistry
Jian, J., Zhang, C., Ke, L., Hayes, N., Zhu, Y., Qiu, H., Zhang, B., Zhou, T., & Wei, G. (2025) · Chemical Science · doi:10.1039/d5sc00270b · Cited by 76
Self-citationADCdb: the database of antibody–drug conjugates
Shen, L., Sun, X., Chen, Z., Guo, Y., Shen, Z., Song, Y., Xin, W., Ding, H., Ma, X., Xu, W., et al. (2023) · Nucleic Acids Research · doi:10.1093/nar/gkad831 · Cited by 81
“The efficacy of any studied ADC could be determined by different perspectives: whether the antibody could accurately bind to the antigen in lesion site ( 52–58 ) and whether the payload could exert effective therapeutic effects in disease tissue ( 59 ,60 ).”
BackgroundPersistent Laplacian projected Omicron BA.4 and BA.5 to become new dominating variants
Chen, J., Qiu, Y., Wang, R., & Wei, G. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.106262 · Cited by 70
“In April 2021, the TopNetTree model predicted a list of 31 RBD antibody-escape mutations, including W353R, I401N, Y449D, Y449S, P491R, P491L, Q493P, etc. [50].”
BackgroundMethodInfluentialSelf-citationRevealing the Threat of Emerging SARS-CoV-2 Mutations to Antibody Therapies
Chen, J., Gao, K., Wang, R., & Wei, G. (2021) · Journal of Molecular Biology · doi:10.1016/j.jmb.2021.167155 · Cited by 72
Self-citationA Review of Deep Learning Methods for Antibodies
Graves, J., Byerly, J., Priego, E., Makkapati, N., Parish, S., Medellin, B. P., & Berrondo, M. (2020) · Antibodies · doi:10.3390/antib9020012 · Cited by 76
“TopNetTree The need to treat antibody–antigen interfaces as special cases of protein–protein interfaces is further reinforced by the analysis of a deep learning method termed TopNetTree [41].”
MethodInfluentialDe novo generation of SARS-CoV-2 antibody CDRH3 with a pre-trained generative large language model
He, H., He, B., Guan, L., Zhao, Y., Jiang, F., Chen, G., Zhu, Q.-G., Chen, C. Y., Li, T., & Yao, J. (2024) · Nature Communications · doi:10.1038/s41467-024-50903-y · Cited by 73
“Previous AI methods have utilized antibody structural information to predict the antigen-antibody binding af fi nity 27 – 29 .”
Zhou, Y., Myung, Y., Rodrigues, C. H. M., & Ascher, D. B. (2024) · Nucleic Acids Research · doi:10.1093/nar/gkae412 · Cited by 65
“DDMut-PPI achieved competitive performance against the top methods ( 12 , 24 , 26 , 34 , 35 , 38–44 ) across these diverse datasets (Figure 2 B and Supplementary Tables S3 –S5 ).”
MethodCluster learning-assisted directed evolution
Qiu, Y., Hu, J., & Wei, G. (2021) · Nature Computational Science · doi:10.1038/s43588-021-00168-y · Cited by 63
Self-citationOmicron BA.2 (B.1.1.529.2): high potential to becoming the next dominating variantPreprint
Chen, J., & Wei, G. (2022) · Research Square · doi:10.21203/rs.3.rs-1362445/v1 · Cited by 69
“2, which is the same as reported in an initial study [12].”
Self-citationArtificial intelligence-driven biomedical genomics
Guo, K., Wu, M., Soo, Z., Yang, Y. Y., Zhang, Y., Zhang, Q., Lin, H., Grosser, M., Venter, D. J., Zhang, G., et al. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.110937 · Cited by 63
Persistent spectral theory-guided protein engineering
(2023) · doi:10.1038/s43588-022-00394-y · Cited by 62
“Persistent homology, a main workhorse of TDA, was integrated with machine learning models to reduce the structural complexity in drug designs [27], and protein-protein interactions [28].”
BackgroundSelf-citationOpen-Source Machine Learning in Computational Chemistry
Hagg, A., & Kirschner, K. N. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c00643 · Cited by 65
Hummer, A. M., Schneider, C., Chinery, L., & Deane, C. M. (2025) · Nature Computational Science · doi:10.1038/s43588-025-00823-8 · Cited by 62
“This value (~0.64) was lower than the Pearson’s methods are trained on features derived from antibody–antigen complex structures, such as binding surface area, interatomic interactions and energy-based terms 14,17,18 .”
InfluentialDeep mutational scanning for therapeutic antibody engineering
Hanning, K. R., Minot, M., Warrender, A. K., Kelton, W., & Reddy, S. T. (2021) · Trends in Pharmacological Sciences · doi:10.1016/j.tips.2021.11.010 · Cited by 50
Accelerating therapeutic protein design with computational approaches toward the clinical stage
Chen, Z., Wang, X., Chen, X., Huang, J., Wang, C., Wang, J., & Wang, Z. (2023) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2023.04.027 · Cited by 49
Toward real-world automated antibody design with combinatorial Bayesian optimization
Khan, A. A., Cowen-Rivers, A. I., Grosnit, A., Deik, D.-G.-X., Robert, P. A., Greiff, V., Smorodina, E., Rawat, P., Akbar, R., Dreczkowski, K., et al. (2023) · Cell Reports Methods · doi:10.1016/j.crmeth.2022.100374 · Cited by 71
Wee, J., & Xia, K. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac024 · Cited by 48
Sun, Y., Zhuang, Z., Xu, H., Qin, W., & Feng, M.-J. (2021) · Journal of Manufacturing Systems · doi:10.1016/j.jmsy.2021.04.001 · Cited by 44
Li, G., Pahari, S., Murthy, A. K., Liang, S., Fragoza, R., Yu, H., & Alexov, E. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa761 · Cited by 44
Machine learning methods for protein-protein binding affinity prediction in protein design
Guo, Z., & Yamaguchi, R. (2022) · Frontiers in Bioinformatics · doi:10.3389/fbinf.2022.1065703 · Cited by 41
“developed a topology-based network to capture the geometric and topological pattern of the complex efficiently (Wang et al., 2020).”
BackgroundImplications of disease-related mutations at protein–protein interfaces
Xiong, D., Lee, D., Li, L., Zhao, Q., & Yu, H. (2021) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2021.11.012 · Cited by 40
Review of COVID-19 Antibody Therapies
Chen, J., Gao, K., Wang, R., Nguyen, D. D., & Wei, G. (2020) · Annual Review of Biophysics · doi:10.1146/annurev-biophys-062920-063711 · Cited by 37
“Topology-Based Network Tree Model for Protein–Protein Interaction Binding Affinity Changes Upon Mutation In this section, we describe the TopNetTree model, which predicts binding affinity changes following mutation ΔΔG for PPIs (77).”
InfluentialSelf-citationArtificial intelligence based methods for hot spot prediction
Ovek, D., Abali, Z., Zeylan, M. E., Keskin, Ö., Gürsoy, A., & Tunçbağ, N. (2021) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2021.11.003 · Cited by 36
Molecular geometric deep learning
Shen, C., Luo, J., & Xia, K. (2023) · Cell Reports Methods · doi:10.1016/j.crmeth.2023.100621 · Cited by 35
“Recently, non-covalent-interaction-based molecular descriptors have achieved great performance in the prediction of binding affinities of protein-ligand and protein-protein [24, 25], demonstrating potential new molecular graph representations beyond the de facto standard of covalent-bond-based graphs.”
BackgroundKim, D. N., McNaughton, A. D., & Kumar, N. (2024) · Bioengineering · doi:10.3390/bioengineering11020185 · Cited by 34
“Many researchers proposed that the fundamental principles governing PPI prediction can be adapted to predict Ab–Ag interactions as well [50].”
BackgroundSVSBI: sequence-based virtual screening of biomolecular interactions
Shen, L., Feng, H., Qiu, Y., & Wei, G. (2023) · Communications Biology · doi:10.1038/s42003-023-04866-3 · Cited by 32
Self-citationJin, R., Ye, Q., Wang, J., Cao, Z., Jiang, D., Wang, T., Kang, Y., Xu, W., Hsieh, C., & Hou, T. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae304 · Cited by 33
“The constrained applicability of physics-based models lies in their primary focus on elucidating interaction mechanisms, rendering them less suitable for guiding antibody optimization routes and screenings [25].”
BackgroundMethodZhou, Z., Yin, Y., Han, H., Jia, Y., Koh, J. H., Kong, A. W.-K., & Mu, Y. (2024) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.4c01850 · Cited by 30
A Review of Topological Data Analysis and Topological Deep Learning in Molecular Sciences
Wee, J., & Jiang, J. W. (2025) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.5c02266 · Cited by 36
“This topological perspective enhances predictive power, improves generalization across diverse chemical scaffolds, and facilitates the identification of novel drug candidates beyond conventional chemical similarity boundaries.”
Vandaele, R., Mukherjee, P., Selby, H. M., Shah, R. P., & Gevaert, O. (2022) · Patterns · doi:10.1016/j.patter.2022.100657 · Cited by 26
“This includes tasks such as predicting protein–protein interaction binding affinity changes [6], biomedical network classification [7, 8], survival prediction of cancer patients [9, 10], and skin lesion segmentation [11].”
BackgroundMethodQuantification of biases in predictions of protein–protein binding affinity changes upon mutations
Tsishyn, M., Pucci, F., & Rooman, M. J. (2023) · Briefings in Bioinformatics · doi:10.1093/bib/bbad491 · Cited by 33
“NetTree [20] is a deep learning method based on convolutional neural networks and algebraic topology features.”
BackgroundMethodAspects of topological approaches for data science
Grbić, J., Wu, J., Xia, K., & Wei, G. (2022) · Foundations of Data Science · doi:10.3934/fods.2022002 · Cited by 24
“TDAbased machine learning models [41, 112] have been used in various areas, including image analysis [9, 80, 113, 96, 109, 68], shape analysis [22, 146, 93, 79, 144], timeseries data analysis [125, 6, 145, 134, 131], computational biology [110, 51, 30, 33], noise data [108], sphere packing [118], language analysis [14…”
MethodSelf-citationCLADE 2.0: Evolution-Driven Cluster Learning-Assisted Directed Evolution
Qiu, Y., & Wei, G. (2022) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.2c01046 · Cited by 23
Self-citationMohseni Behbahani, Y., Laine, E., & Carbone, A. (2023) · Bioinformatics · doi:10.1093/bioinformatics/btad231 · Cited by 23
“Finally, we considered two recent deep learning-based approaches, namely GraphPPI [42] and TopNetTree [73].”
MethodInfluentialYu, G., Zhao, Q., Bi, X., & Wang, J. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae232 · Cited by 23
Jiang, Y., Quan, L., Li, K., Li, Y., Zhou, Y., Wu, T., & Lyu, Q. (2023) · IEEE Transactions on Computational Biology and Bioinformatics · doi:10.1109/tcbb.2022.3233627 · Cited by 21
Riahi, S., Lee, J., Wei, S., Cost, R., Masiero, A., Prades, C., Olfati-Saber, R., Wendt, M., Park, A., Qiu, Y., et al. (2021) · Antibody Therapeutics · doi:10.1093/abt/tbab011 · Cited by 19
Liu, X., Feng, H., Wu, J., & Xia, K. (2022) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.2c00580 · Cited by 19
Yue, Y., Li, S., Wang, L., Liu, H., Tong, H. H. Y., & He, S. (2023) · Briefings in Bioinformatics · doi:10.1093/bib/bbad310 · Cited by 19
Wee, J., Chen, J., Xia, K., & Wei, G. (2024) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2024.107918 · Cited by 20
“…homology was integrated with machine learning for the classification of proteins in 2015 [15], which was one the first integrations of TDA and machine learning, and the predictions of mutation-induced protein stability changes [16, 17] and protein-protein binding free energy changes [18, 19].”
MethodSelf-citationDecoding the functional impact of the cancer genome through protein–protein interactions
Fu, H., Mo, X., & Ivanov, A. A. (2025) · Nature reviews. Cancer · doi:10.1038/s41568-024-00784-6 · Cited by 22
“Such methods have been successfully used for PPI predictions 163,164 .”
PPB-Affinity: Protein-Protein Binding Affinity dataset for AI-based protein drug discovery
Liu, H., Chen, P., Zhai, X., Huo, K.-G., Zhou, S., Han, L., & Fan, G. (2024) · Scientific Data · doi:10.1038/s41597-024-03997-4 · Cited by 30
Villalobos-Alva, J., Ochoa-Toledo, L., Villalobos-Alva, M. J., Aliseda, A., Escamirosa, F. P., Altamirano-Bustamante, N. F., Ochoa-Fernández, F., Zamora-Solís, R., Villalobos-Alva, S., Revilla, M., et al. (2022) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2022.788300 · Cited by 19
Lim, H., Jeon, H.-N., Lim, S., Jang, Y., Kim, T., Cho, H., Pan, J., & No, K. T. (2022) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2022.01.027 · Cited by 17
“Our prediction models in the two datasets outperformed the general-purpose prediction models in protein expression [52] and binding affinity [53–56] and outperformed the specific prediction models for binding affinity in SARS-CoV-2 [57,58].”
BackgroundResultInfluentialEstimating Absolute Protein–Protein Binding Free Energies by a Super Learner Model
Chaves, E. J. F., Sartori, J., Santos, W. M., da Cruz, C. H. B., Mhrous, E. N., Nacimento-Filho, M. F., Ferraz, M. V. F., & Lins, R. D. (2025) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.4c01641 · Cited by 17
Topological deep learning based deep mutational scanning
Chen, J., Woldring, D., Huang, F., Huang, X., & Wei, G. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.107258 · Cited by 15
Self-citationWang, B., Zhou, X., Wang, Y., Gao, Y., Nakanishi, H., Fujita, M., & Li, Z. (2024) · International Journal of Biological Macromolecules · doi:10.1016/j.ijbiomac.2024.134481 · Cited by 16
Han, H., Jiang, B., Shi, Y., Jiang, P., Zhang, H., Shang, C., Sun, Y., Li, Y., & Xiang, D. (2023) · Ecological Indicators · doi:10.1016/j.ecolind.2023.110809 · Cited by 16
General Theory of Specific Binding: Insights from a Genetic-Mechano-Chemical Protein Model
McBride, J. M., Eckmann, J., & Tlusty, T. (2022) · Molecular Biology and Evolution · doi:10.1093/molbev/msac217 · Cited by 15
High Bandwidth Utilization DWBA Algorithm for Upstream Channel in NG-EPON
Hui, J., Gan, C., Wu, L., & Xu, Z. (2022) · IEEE Access · doi:10.1109/access.2022.3207170 · Cited by 14
“The simulation is based on a typical tree network topology [20], [21], [22].”
MethodSequence-based deep learning antibody design for in silico antibody affinity maturationPreprint
Kang, Y., Leng, D., Guo, J., & Pan, L. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.03724 · Cited by 14
“[5] studied the prediction of interacting contacts and performed binding prediction using surface-based geometric features; [4] employed topologybased network tree for binding affinity change prediction based on complexes’ 3D structures.”
Mayer-Homology Learning Prediction of Protein-Ligand Binding Affinities
Feng, H., Shen, L., Liu, J., & Wei, G.-W. (2024) · Journal of Computational Biophysics and Chemistry · doi:10.1142/s2737416524500613 · Cited by 14
“Its integration with machine learning algorithms has led to significant advancements in various stages of drug design such as predicting protein–ligand binding affinity [18] and changes in protein stability upon mutation [24].”
BackgroundSelf-citationWee, J., & Wei, G.-W. (2024) · PubMed · doi:10.48550/arxiv.2406.03979 · Cited by 13
“One of the very successful approaches for mutation-induced protein-protein BFE change predictions is the topology-based network tree (TopNetTree) [12], which is based on topological deep learning (TDL), introduced in 2017 [13].”
BackgroundMethodInfluentialSelf-citationMuToN Quantifies Binding Affinity Changes upon Protein Mutations by Geometric Deep Learning
Li, P., & Liu, Z. (2024) · Advanced Science · doi:10.1002/advs.202402918 · Cited by 11
Rana, M. M., & Nguyen, D. D. (2023) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.3c02679 · Cited by 11
Yue, Y., Li, S., Cheng, Y., Wang, L., Hou, T., Zhu, Z., & He, S. (2024) · Nature Communications · doi:10.1038/s41467-024-53583-w · Cited by 11
“Speci-fi cally, based on a line graph-enhanced edge message passing mechanism 53 to model the inter-edge positional relationships, the additional structural information can be injected into the node representations for more effective protein geometric interaction modeling (the corresponding equations are provided in Su…”
InfluentialRapid response to fast viral evolution using AlphaFold 3-assisted topological deep learning
Wee, J., & Wei, G. (2025) · Virus Evolution · doi:10.1093/ve/veaf026 · Cited by 12
“Recently, persistent Laplacian (PL) has been introduced to address these limitations of existing TDA methods Wang et al. (2020).”
InfluentialSelf-citationOn the linkage of thermodynamics and pathogenicity
Pandey, P., Ghimire, S., Wu, B., & Alexov, E. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102572 · Cited by 11
AI-driven prediction of SARS-CoV-2 variant binding trends from atomistic simulations
Capponi, S., Wang, S., Navarro, E. J., & Bianco, S. (2021) · The European Physical Journal E · doi:10.1140/epje/s10189-021-00119-5 · Cited by 12
Circuit Topology Approach for the Comparative Analysis of Intrinsically Disordered Proteins
Scalvini, B., Sheikhhassani, V., van de Brug, N., Heling, L. W. H. J., Schmit, J. D., & Mashaghi, A. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c00391 · Cited by 10
Circuit topology predicts pathogenicity of missense mutations
Woodard, J. C., Iqbal, S., & Mashaghi, A. (2022) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.26342 · Cited by 10
Persistent Tor-algebra for protein–protein interaction analysis
Liu, X., Feng, H., Lü, Z., & Xia, K. (2023) · Briefings in Bioinformatics · doi:10.1093/bib/bbad046 · Cited by 11
Han, H., Jiang, B., Xiang, D., Shi, Y., Liu, S., Shang, C., Zhao, X., Zhang, H., & Sun, Y. (2024) · Ecological Indicators · doi:10.1016/j.ecolind.2023.111526 · Cited by 9
Xu, Q., Li, Y., & Xie, M. (2024) · Water Air & Soil Pollution · doi:10.1007/s11270-024-07457-y · Cited by 9
“…through convolutional neural network (CNN), which has the powerful ability of image processing and has been widely used in biology, ecology, and medicine, such as protein–protein-binding (Wang et al., 2020), cancer diagnosis (Tolkach et al., 2020), face recognition (Hill et al., 2019), etc.”
Zhang, Y., Dong, M., Deng, J., Wu, J., Zhao, Q., Gao, X., & Xiong, D. (2024) · Communications Biology · doi:10.1038/s42003-024-07066-9 · Cited by 9
“In recent years, the deep learning-based methods 15,24 – 28 , therefore, have rapidly emerged to pave avenues for the prediction of ΔΔ G s in PPIs.”
InfluentialPersistent Topological Laplacian Analysis of SARS-CoV-2 Variants
Wei, X., Chen, J., & Wei, G. (2023) · Journal of Computational Biophysics and Chemistry · doi:10.1142/s2737416523500278 · Cited by 9
Self-citationGrowing ecosystem of deep learning methods for modeling protein–protein interactions
Rogers, J. R., Nikolényi, G., & AlQuraishi, M. (2023) · Protein Engineering Design and Selection · doi:10.1093/protein/gzad023 · Cited by 12
Mathematical artificial intelligence design of mutation-proof COVID-19 monoclonal antibodies
Chen, J., & Wei, G. (2022) · Communications in Information and Systems · doi:10.4310/cis.2022.v22.n3.a3 · Cited by 9
Self-citationAntBO: Towards Real-World Automated Antibody Design with Combinatorial Bayesian OptimisationPreprint
Khan, A., Cowen-Rivers, A. I., Deik, D.-G.-X., Grosnit, A., Robert, P. A., Greiff, V., Smorodina, E., Rawat, P., Akbar, R., Dreczkowski, K., et al. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4115860 · Cited by 9
Lombardi, G., & Carbone, A. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.08.24.609515 · Cited by 10
“This can be addressed either by deliberatly overlooking the asymmetry, resulting in a predisposition for positive predictions, or by augmenting the database with inverse mutations, as in (Wang et al. , 2020), coupled with incorporating a distinct term in the loss function to attain symmetry.”
MethodSun, X., Yang, S., Wu, Z., Su, J., Hu, F., Chang, F., & Li, C. (2024) · Structure · doi:10.1016/j.str.2024.02.016 · Cited by 8
Bi, J., Wee, J., Liu, X., Qu, C., Wang, G., & Xia, K. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c00621 · Cited by 8
Preventing future zoonosis: SARS-CoV-2 mutations enhance human–animal cross-transmission
Wee, J., Chen, J., & Wei, G. (2024) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2024.109101 · Cited by 7
Self-citationProtein Engineering for Enhancing Electron Transfer in P450‐Mediated Catalysis
Li, Y., Yu, H., & Ye, L. (2025) · Biotechnology and Bioengineering · doi:10.1002/bit.29023 · Cited by 9
“An increasing number of protein – protein interaction prediction tools (Rodrigues et al. 2021; Wang et al. 2020; Zhou, Yin, et al. 2024) have been developed, providing a feasible approach for modifying the interface between P450s and RPs. Methods like bimolecular fluorescence complementation…”
Multi-Cover Persistence (MCP)-based machine learning for polymer property prediction
Zhang, Y., Shen, C., & Xia, K. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae465 · Cited by 7
“…prediction [18,22–24], protein stability change upon mutation prediction [20,25],toxicity prediction [26–28],solvation free energy prediction [ 29,30], partition coefficient and aqueous solubility [31], binding pocket detection [32], protein mutation analysis [33–35], and drug discovery [36].”
BackgroundRidha, F., & Gromiha, M. M. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae598 · Cited by 7
Metadata
- Authors
- Wang, M., Cang, Z., & Wei, G.-W.
- Affiliations
- Michigan State University (US)
- Corresponding author
- Wei, Guo-Wei
- Venue
- Nature Machine Intelligence
- Published
- 2020-02-14
- Volume
- 2
- Issue
- 2
- Pages
- 116-123
- ISSN
- 2522-5839 (ISSN-L)
- PMC
- PMC7223817
- OpenAlex
- W3005822506
- Semantic Scholar
- 544ee14a801abef4019e1962a38e9a1f936ef01a
- Cited by
- 202
- References
- 54
- Citations by year
- 2026: 19 · 2025: 41 · 2024: 39 · 2023: 41 · 2022: 32 · 2021: 22 · 2020: 7
- FWCI
- 8.27
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Licence
- Locations
- Nature Machine Intelligence (publishedVersion)
- PubMed Central (publishedVersion)
- PubMed (publishedVersion)
- Funders
- National Science Foundation (DMS-1721024, DMS 1761320, DMS-1761320, IIS1900473)
- U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences (GM126189)
- Bristol-Myers Squibb
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Topological and Geometric Data Analysis
- Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
- Keywords
- protein-protein interaction, ΔΔG, mutation effects, persistent homology, algebraic topology, protein-protein complex, convolutional neural networks, gradient boosting decision tree, topological invariants
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
