Leveraging Uncertainty in Machine Learning Accelerates Biological Discovery and Design
Open accessPublished version of a preprint
Abstract
Machine learning that generates biological hypotheses has transformative potential, but most learning algorithms are susceptible to pathological failure when exploring regimes beyond the training data distribution. A solution to address this issue is to quantify prediction uncertainty so that algorithms can gracefully handle novel phenomena that confound standard methods. Here, we demonstrate the broad utility of robust uncertainty prediction in biological discovery. By leveraging Gaussian process-based uncertainty prediction on modern pre-trained features, we train a model on just 72 compounds to make predictions over a 10,833-compound library, identifying and experimentally validating compounds with nanomolar affinity for diverse kinases and whole-cell growth inhibition of Mycobacterium tuberculosis. Uncertainty facilitates a tight iterative loop between computation and experimentation and generalizes across biological domains as diverse as protein engineering and single-cell transcriptomics. More broadly, our work demonstrates that uncertainty should play a key role in the increasing adoption of machine learning algorithms into the experimental lifecycle.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
In an AI/ML study based on kinase profiling data, considering confidence intervals remarkably increased the relevance of selected virtual hits.
“Indeed, in an AI/ML study based on kinase profiling data, considering confidence intervals remarkably increased the relevance of selected virtual hits.280”
From Advancing Targeted Protein Degradation via Multiomics Profiling and Artificial Intelligence (2023) · cites this paper as [280]
Language models learn summary representations that can be used to capture sequence relationships and link sequence to function via transfer learning.
“They also learn summary representations, powerful features that can be used to better capture sequence relationships and link sequence to function via transfer learning 27,28,32-35.”
From Learning the protein language: Evolution, structure, and function (2021) · cites this paper as [35]
Hie et al. presented a GP-based method for guiding experimental drug design informed by deep protein embeddings.
“Hie et al, presented a GP-based method for guiding experimental drug design informed by deep protein embeddings 35.”
From Learning the protein language: Evolution, structure, and function (2021) · cites this paper as [35]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing
Benjamini, Y., & Hochberg, Y. (1995) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1995.tb02031.x · Cited by 109,482 · see citing works
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · see citing works
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126 · see citing works
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 · see citing works
Trott, O., & Olson, A. J. (2010) · Journal of Computational Chemistry · doi:10.1002/jcc.21334 · Cited by 38,412 · cited 2× · see citing works
“…prediction of true biochemical affinity (Palacio-Rodrı ´guez et al., 2019), we used six scoring functions to compare generated designs selected with and without uncertainty (Koes et al., 2013; Trott and Olson, 2010; Quiroga and Villarreal, 2016; Ruiz-Carmona et al., 2014; Zhao and Huang, 2011).”
MethodInfluential(2008) · Journal of Machine Learning Research · Cited by 33,051 · see citing works
AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility
Morris, G. M., Huey, R., Lindstrom, W., Sanner, M. F., Belew, R. K., Goodsell, D. S., & Olson, A. J. (2009) · Journal of Computational Chemistry · doi:10.1002/jcc.21256 · Cited by 25,745 · see citing works
Fast unfolding of communities in large networks
Blondel, V. D., Guillaume, J.-L., Lambiotte, R., & Lefebvre, E. (2008) · Journal of Statistical Mechanics: Theory and Experiment · doi:10.1088/1742-5468/2008/10/p10008 · Cited by 21,424 · see citing works
Auto-Encoding Variational Bayes
Kingma, D. P., & Welling, M. (2013) · arXiv · doi:10.48550/arxiv.1312.6114 · Cited by 20,067 · see citing works
Integrating single-cell transcriptomic data across different conditions, technologies, and species
Butler, A., Hoffman, P., Smibert, P., Papalexi, E., & Satija, R. (2018) · Nature Biotechnology · doi:10.1038/nbt.4096 · Cited by 13,174 · see citing works
Gaussian Processes for Machine Learning
Rasmussen, C. E., & Williams, C. K. I. (2005) · doi:10.7551/mitpress/3206.001.0001 · Cited by 11,705 · cited 5× · see citing works
“GPs are prime candidates for machine learning-based hypothesis generation since they naturally quantify prediction uncertainty ( Rasmussen and Williams, 2005), are highly sample effi-cient (Grande et al., 2014), and can readily incorporate a rich set of features like those obtained by pretraining.”
InfluentialOpen Babel: An open chemical toolbox
O'Boyle, N. M., Banck, M. S., James, C. A. J., Morley, C., Vandermeersch, T., & Hutchison, G. R. (2011) · Journal of Cheminformatics · doi:10.1186/1758-2946-3-33 · Cited by 10,705 · cited 1× · see citing works
“Structure files were preprocessed to be in compatible file formats using AutoDockTools version 1.5.6 and Open Babel version 2.3.2 (O’Boyle et al., 2011).”
MethodInfluentialSCANPY: large-scale single-cell gene expression data analysis
Wolf, F. A., Angerer, P., & Theis, F. J. (2018) · Genome Biology · doi:10.1186/s13059-017-1382-0 · Cited by 8,881 · see citing works
UniProt: a worldwide hub of protein knowledge
The UniProt Consortium (2019) · Nucleic Acids Research · doi:10.1093/nar/gky1049 · Cited by 7,698 · cited 1× · see citing works
“Compound structures were obtained from the ZINC database (Irwin and Shoichet, 2005) and kinase amino acid sequences were obtained from the UniProt database (UniProt Consortium, 2019).”
MethodThe Logic of Scientific Discovery.
Hutchison, T. W., & Popper, K. R. (1959) · Economica · doi:10.2307/2550489 · Cited by 7,628 · see citing works
Extended-Connectivity Fingerprints
Rogers, D. O. W., & Hahn, M. (2010) · Journal of Chemical Information and Modeling · doi:10.1021/ci100050t · Cited by 7,597 · see citing works
Weininger, D. (1988) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci00057a005 · Cited by 7,436 · cited 1× · see citing works
“For compound features, we provided the model with chemical SMILE strings that the model transforms into a graph convolutional representation (Weininger, 1988); for kinase features, we use the protein contact maps provided by the original study.”
MethodBase-Calling of Automated Sequencer Traces Using Phred. I. Accuracy AssessmentPreprint
Ewing, B., Hillier, L., Wendl, M. C., & Green, P. (1998) · Genome Research · doi:10.1101/gr.8.3.175 · Cited by 6,993 · cited 1× · see citing works
“…et al., 2017; Neal, 2012), in the biological setting, many machine learning studies do not consider uncertainty or are limited to specific tasks or in-silico validations (Cortes-Ciriano et al., 2015; Ewing et al., 1998; Norinder et al., 2014; Sverchkov and Craven, 2017; Zeng and Gifford, 2019).”
BackgroundNWChem: A comprehensive and scalable open-source solution for large scale molecular simulations
Valiev, M., Bylaska, E. J., Govind, N., Kowalski, K., Straatsma, T. P., van Dam, H. J. J., Wang, D., Nieplocha, J., Aprà, E., Windus, T. L., et al. (2010) · Computer Physics Communications · doi:10.1016/j.cpc.2010.04.018 · Cited by 5,794 · see citing works
Bayesian Learning for Neural Networks
Neal, R. M. (1996) · Lecture Notes in Statistics · doi:10.1007/978-1-4612-0745-0 · Cited by 4,276 · cited 14× · see citing works
“We train an MLP regressor with the same architecture described above (two hidden layers with 200 neurons per layer and ReLU non linearities) but with a unit-variance Gaussian prior on each weight and bias entry (Neal, 2012).”
BackgroundMethodInfluentialZINC − A Free Database of Commercially Available Compounds for Virtual Screening
Irwin, J. J., & Shoichet, B. K. (2005) · Journal of Chemical Information and Modeling · doi:10.1021/ci049714+ · Cited by 4,058 · cited 4× · see citing works
“We used the ZINC database (Irwin and Shoichet, 2005), an online repository of chemical compounds with associated metadata for each compound that includes their structure and their commercial availability.”
MethodInfluentialSimple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
Lakshminarayanan, B., Pritzel, A., & Blundell, C. (2016) · arXiv · doi:10.48550/arxiv.1612.01474 · Cited by 3,666 · cited 12× · see citing works
“GP, MLP + GP, BMLP, GMLPE, MLP, and OLS regressors were each trained on 1,115 unique sequences containing at most one mutation to wild-type avGFP (UniProt: P42212).”
BackgroundMethodInfluentialOn calibration of modern neural networks
Guo, C., Pleiss, G., Sun, Y., & Weinberger, K. Q. (2017) · arXiv · doi:10.48550/arxiv.1706.04599 · Cited by 3,435 · cited 1× · see citing works
“…expert, a standard machine learning algo-rithm has no corresponding notion of uncertainty, potentially leading to biased, overconfident, or pathological predictions (Figure 1A) (Amodei et al., 2016; Chen et al., 2018; Guo et al., 2017; Lakshminarayanan et al., 2017; Neal, 2012; Nguyen et al., 2015).”
BackgroundPython for scientific computing
Oliphant, T. E. (2007) · Computing in Science & Engineering · doi:10.1109/mcse.2007.58 · Cited by 3,303 · cited 1× · see citing works
“We use the scientific Python toolkit, including the scipy (version 1.3.1) and numpy (version 1.17.2) Python packages (Oliphant, 2007), to compute the statistical tests described in the manuscript, including Pearson correlation, Spearman correlation, Welch’s unequal variance t-test, and associated P…”
MethodInfluentialDeep neural networks are easily fooled: High confidence predictions for unrecognizable images
Nguyen, A., Yosinski, J., & Clune, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298640 · Cited by 3,258 · cited 1× · see citing works
“While notions of biochemical ‘‘similarity’’ or ‘‘uncertainty’’ might be obvious to a human expert, a standard machine learning algorithm has no corresponding notion of uncertainty, potentially leading to biased, overconfident, or pathological predictions (Figure 1A) (Amodei et al., 2016; Chen et al., 2018; Guo et al.,…”
BackgroundUnderstanding Machine Learning: From Theory to Algorithms
Shalev-Shwartz, S., & Ben-David, S. (2014) · doi:10.1017/cbo9781107298019 · Cited by 3,115 · see citing works
What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
Kendall, A., & Gal, Y. (2017) · arXiv · doi:10.48550/arxiv.1703.04977 · Cited by 3,088 · cited 1× · see citing works
“While uncertainty is gradually becoming recognized as a critical property in learning algorithms (Kendall and Gal, 2017; Lakshminarayanan et al., 2017; Neal, 2012), in the biological setting, many machine learning studies do not consider uncertainty or are limited to specific tasks or in-silico…”
BackgroundPan-cancer analysis of whole genomes
The ICGC/TCGA Pan-Cancer Analysis of Whole Genomes Consortium, Aaltonen, L. A., Abascal, F., Abeshouse, A., Aburatani, H., Adams, D. J., Agrawal, N., Ahn, K. S., Ahn, S.-M., Aikata, H., et al. (2020) · Nature · doi:10.1038/s41586-020-1969-6 · Cited by 3,064 · see citing works
Jacomy, M., Venturini, T., Heymann, S., & Bastian, M. (2014) · PLoS ONE · doi:10.1371/journal.pone.0098679 · Cited by 2,909 · see citing works
(2015) · Cited by 2,857 · see citing works
<i>The Logic of Scientific Discovery</i>
Popper, K. R., & Weiss, G. (1959) · Physics Today · doi:10.1063/1.3060577 · Cited by 2,717 · see citing works
FACS-optimized mutants of the green fluorescent protein (GFP)
Cormack, B. P., Valdivia, R. H., & Falkow, S. (1996) · Gene · doi:10.1016/0378-1119(95)00685-0 · Cited by 2,621 · cited 1× · see citing works
“These results suggest how predictive (and robust) algorithms could help reduce the complexity of mutagenesis experiments traditionally used to optimize new protein designs (Cormack et al., 1996; Sarkisyan et al., 2016).”
BackgroundA Deep Learning Approach to Antibiotic Discovery
Stokes, J. M., Yang, K., Swanson, K., Jin, W., Cubillos-Ruiz, A., Donghia, N. M., MacNair, C. R., French, S., Carfrae, L. A., Bloom-Ackermann, Z., et al. (2020) · Cell · doi:10.1016/j.cell.2020.01.021 · Cited by 2,463 · cited 1× · see citing works
“Machine learning algorithms offer a way to translate existing data into actionable biological hypotheses (Bogard et al., 2019; King et al., 2004; LeCun et al., 2015; Stokes et al., 2020; Tarca et al., 2007; Yang et al., 2019).”
BackgroundThe Logic of Scientific Discovery.
Rescher, N., Popper, K. R., & Hanson, N. R. (1960) · Philosophy and Phenomenological Research · doi:10.2307/2104331 · Cited by 2,455 · see citing works
Comprehensive analysis of kinase inhibitor selectivity
Davis, M. I., Hunt, J. P., Herrgard, S., Ciceri, P., Wodicka, L. M., Pallares, G., Hocker, M., Treiber, D. K., & Zarrinkar, P. P. (2011) · Nature Biotechnology · doi:10.1038/nbt.1990 · Cited by 2,409 · cited 7× · see citing works
“We train our models with information from just 72 kinase inhibitors (Davis et al., 2011), perform an in-silico screen of an unbiased 10,833-compound chemical library (Irwin and Shoichet, 2005), and experimentally validate our machine-generated hypotheses with in-vitro binding assays.”
BackgroundMethodInfluentialWhy Does Unsupervised Pre-training Help Deep Learning?
Erhan, D., Bengio, Y., Courville, A. C., Manzagol, P.-A., Vincent, P., & Samy Bengio, A. (2010) · Cited by 2,180 · see citing works
Proceedings of the 19th International Conference on Artificial Intelligence and Statistics 2016
Murray, I., & Graham, M. (2016) · Cited by 2,079 · see citing works
Recovering Gene Interactions from Single-Cell Data Using Data Diffusion
van Dijk, D., Sharma, R., Nainys, J., Yim, K., Kathail, P., Carr, A. J., Burdziak, C., Moon, K. R., Chaffer, C. L., Pattabiraman, D., et al. (2018) · Cell · doi:10.1016/j.cell.2018.05.061 · Cited by 1,829 · see citing works
Junction Tree Variational Autoencoder for Molecular Graph GenerationPreprint
Jin, W.-G., Barzilay, R., & Jaakkola, T. (2018) · International Conference on Machine Learning · doi:10.48550/arxiv.1802.04364 · Cited by 1,783 · cited 10× · see citing works
“Our generation strategy was based on sampling from the latent space of a variational autoencoder (VAE) (Kingma and Welling, 2014), with an architecture optimized for chemical structures (JTNN-VAE) (Jin et al., 2018) (STARMethods).”
MethodInfluentialDeepDTA: deep drug–target binding affinity prediction
Öztürk, H., Özgür, A., & Özkırımlı, E. (2018) · Bioinformatics · doi:10.1093/bioinformatics/bty593 · Cited by 1,735 · see citing works
Deep learning enables rapid identification of potent DDR1 kinase inhibitors
Zhavoronkov, A., Ivanenkov, Y. A., Aliper, A., Veselov, M. S., Aladinskiy, V. A., Aladinskaya, A. V., Terentiev, V. A., Polykovskiy, D. A., Kuznetsov, M. D., Asadulaev, A., et al. (2019) · Nature Biotechnology · doi:10.1038/s41587-019-0224-x · Cited by 1,573 · see citing works
Learning to detect unseen object classes by between-class attribute transfer
Lampert, C. H., Nickisch, H., & Stefan Harmeling, B. (2009) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvprw.2009.5206594 · Cited by 1,563 · see citing works
GARCÍA-ALONSO, C. R., Pérez, L. M., & Fernández, J. C. F. (2011) · Annals of Operations Research · doi:10.1007/s10479-011-0841-3 · Cited by 1,357 · see citing works
Concrete Problems in AI Safety
Amodei, D., Olah, C., Steinhardt, J., Christiano, P., Schulman, J., & Mané, D. (2016) · arXiv · doi:10.48550/arxiv.1606.06565 · Cited by 1,314 · cited 2× · see citing works
“While notions of biochemical ‘‘similarity’’ or ‘‘uncertainty’’ might be obvious to a human expert, a standard machine learning algorithm has no corresponding notion of uncertainty, potentially leading to biased, overconfident, or pathological predictions (Figure 1A) (Amodei et al., 2016; Chen et al., 2018; Guo et al.,…”
BackgroundRelational learning via collective matrix factorization
Singh, A. P., & Gordon, G. J. (2008) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/1401890.1401969 · Cited by 1,237 · cited 1× · see citing works
“We performed CMF using the compound-kinase Kds as the explicit data matrix and the neural-encoded compound and kinase features as side-information (Singh and Gordon, 2008).”
MethodUsing confidence bounds for exploitation-exploration trade-offs
Auer, P. L. (2003) · Cited by 1,219 · see citing works
Lessons Learned in Empirical Scoring with smina from the CSAR 2011 Benchmarking Exercise
Koes, D. R., Baumgartner, M. P., & Camacho, C. J. (2013) · Journal of Chemical Information and Modeling · doi:10.1021/ci300604z · Cited by 1,150 · cited 2× · see citing works
“…prediction of true biochemical affinity (Palacio-Rodrı ´guez et al., 2019), we used six scoring functions to compare generated designs selected with and without uncertainty (Koes et al., 2013; Trott and Olson, 2010; Quiroga and Villarreal, 2016; Ruiz-Carmona et al., 2014; Zhao and Huang, 2011).”
BackgroundMethodInfluentialMachine-learning-guided directed evolution for protein engineering
Yang, K., Wu, Z., & Arnold, F. H. (2019) · Nature Methods · doi:10.1038/s41592-019-0496-6 · Cited by 1,085 · cited 2× · see citing works
“Machine learning algorithms offer a way to translate existing data into actionable biological hypotheses (Bogard et al., 2019; King et al., 2004; LeCun et al., 2015; Stokes et al., 2020; Tarca et al., 2007; Yang et al., 2019).”
BackgroundMethodRampersad, S. N. (2012) · Sensors · doi:10.3390/s120912347 · Cited by 1,074 · cited 1× · see citing works
“Using the colorimetric, resazurin microtiter assay (alamar blue) (Lougheed et al., 2011; Rampersad, 2012), we determined the minimum inhibitory concentration (MIC) of these compounds as well as rifampicin, a frontline antibiotic for tuberculosis (Furin et al., 2019) (STAR Methods); the MICs for…”
MethodBeing Sensitive to Uncertainty
Arriola, L. M., & Hyman, J. M. (2007) · Computing in Science & Engineering · doi:10.1109/mcse.2007.27 · Cited by 975 · see citing works
Luo, Y., Zhao, X., Zhou, J., Yang, J., Zhang, Y., Kuang, W., Peng, J., Chen, L., & Zeng, J. (2017) · Nature Communications · doi:10.1038/s41467-017-00680-8 · Cited by 924 · cited 2× · see citing works
“, 2018) and collective matrix factorization (CMF), a model often used to recommend shopping items for potential buyers that can also recommend drugs for potential targets (Cobanoglu et al., 2013; Luo et al., 2017; Singh and Gordon, 2008; Zheng et al., 2013); variants of bothmodels have seen extensive use in previous c…”
BackgroundMethodMulti-task Gaussian Process Prediction
Bonilla, E. V., Chai, K. M. A., & Williams, C. K. I. (2007) · Cited by 916 · see citing works
Zero-Shot Learning — The Good, the Bad and the Ugly
Xian, Y., Schiele, B., & Akata, Z. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.328 · Cited by 891 · see citing works
Structure, function, and biogenesis of the cell wall of Mycobacterium tuberculosis
Brennan, P. J. (2003) · Tuberculosis · doi:10.1016/s1472-9792(02)00089-6 · Cited by 888 · cited 2× · see citing works
“11, n = 3 replicate cultures per condition) (Figures 4A and S5A), perhaps due to low cell permeability (Brennan, 2003; Hoffmann et al., 2008).”
BackgroundInfluentialFurin, J. J., Cox, H. S., & Pai, M. (2019) · The Lancet · doi:10.1016/s0140-6736(19)30308-3 · Cited by 876 · see citing works
Efficient integration of heterogeneous single-cell transcriptomes using Scanorama
Hie, B., Bryson, B. D., & Berger, B. (2019) · Nature Biotechnology · doi:10.1038/s41587-019-0113-3 · Cited by 853 · see citing works
Self-citationGPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU AccelerationPreprint
Gardner, J. R., Pleiss, G., Weinberger, K. Q., Bindel, D., & Wilson, A. G. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.11165 · Cited by 721 · cited 1× · see citing works
“For the imputation experiments, we use GPyTorch version 1.0.1 (Gardner et al., 2018) with CUDA acceleration as a more efficient GP implementation than that offered by scikit-learn.”
MethodInfluentialFunctional genomic hypothesis generation and experimentation by a robot scientist
King, R. D., Whelan, K. E., Jones, F. M., Reiser, P. G. K., Bryant, C. H., Muggleton, S. H., Kell, D. B., & Oliver, S. G. (2004) · Nature · doi:10.1038/nature02236 · Cited by 662 · cited 1× · see citing works
“Machine learning algorithms offer a way to translate existing data into actionable biological hypotheses (Bogard et al., 2019; King et al., 2004; LeCun et al., 2015; Stokes et al., 2020; Tarca et al., 2007; Yang et al., 2019).”
BackgroundMachine learning and its applications to biology
Tarca, A. L., Carey, V. J., Chen, X.-W., Romero, R., & Drăghici, S. (2007) · PLoS Computational Biology · doi:10.1371/journal.pcbi.0030116 · Cited by 656 · cited 1× · see citing works
“Machine learning algorithms offer a way to translate existing data into actionable biological hypotheses (Bogard et al., 2019; King et al., 2004; LeCun et al., 2015; Stokes et al., 2020; Tarca et al., 2007; Yang et al., 2019).”
BackgroundrDock: A Fast, Versatile and Open Source Program for Docking Ligands to Proteins and Nucleic Acids
Carmona, S. R., Álvarez-García, D., Foloppe, N., Garmendia-Doval, A. B., Juhos, S., Schmidtke, P., Barril, X., Hubbard, R. E., & Morley, S. (2014) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1003571 · Cited by 645 · see citing works
Hoffmann, C., Leis, A., Niederweis, M., Plitzko, J. M., & Engelhardt, H. (2008) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0709530105 · Cited by 596 · cited 1× · see citing works
“11, n = 3 replicate cultures per condition) (Figures 4A and S5A), perhaps due to low cell permeability (Brennan, 2003; Hoffmann et al., 2008).”
BackgroundLocal fitness landscape of the green fluorescent protein
Sarkisyan, K. S., Bolotin, D. A., Meer, M. V., Usmanova, D. R., Mishin, A. S., Sharonov, G. V., Ivankov, D. N., Bozhanova, N. G., Baranov, M. S., Soylemez, O., et al. (2016) · Nature · doi:10.1038/nature17995 · Cited by 591 · cited 6× · see citing works
“(E) In general, GP-acquired mutations are enriched for surface residues over buried residues as expected (Sarkisyan et al., 2016), with T62 being a notable exception.”
BackgroundMethodResultInfluentialBacon, F. (2000) · Cambridge University Press eBooks · doi:10.1017/cbo9781139164030 · Cited by 569 · see citing works
Inactivation of PI(3)K p110δ breaks regulatory T-cell-mediated immune tolerance to cancer
Ali, K., Soond, D. R., Piñeiro, R., Hagemann, T., Pearce, W. P., Lim, E. L., Bouabe, H., Scudamore, C. L., Hancox, T. C., Maecker, H. L., et al. (2014) · Nature · doi:10.1038/nature13444 · Cited by 530 · cited 5× · see citing works
“Notably, GP acquisition yielded subnanomolar affinities between K252a and IRAK4 (Kd = 0.85 nM) and between PI-3065 and p110d (Kd = 0.36 nM), automating discoveries that previously had been made with massive-scale screens or expert biochemical reasoning (Ali et al., 2014; Ong et al., 2009).”
BackgroundInfluentialInduction of erythroid differentiation in the human leukaemia cell line K562
Andersson, L. C., JOKINEN, M., & Gahmberg, C. G. (1979) · Nature · doi:10.1038/278364a0 · Cited by 475 · cited 1× · see citing works
“Our results suggest that erythroid differentiation of K562, a common in vitromodel of erythropoiesis (Andersson et al., 1979), amplifies the correlation structure that is still present in the precursor state.”
Resultp110δ, a novel phosphoinositide 3-kinase in leukocytes
Vanhaesebroeck, B., Welham, M. J., Kotani, K., Stein, R. C., Warne, P. H., Zvelebil, M., Higashi, K., Volinia, S., Downward, J., & Waterfield, M. D. (1997) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.94.9.4330 · Cited by 434 · cited 1× · see citing works
“…et al., 2007); human c-SRC, a tyrosine kinase and canonical proto-oncogene (Oppermann et al., 1979); human p110 d , a lipid kinase and leukocytic immune regulator (Vanhaesebroeck et al., 1997); and Mtb PknB, a serine-threonine kinase essential to mycobacterial viability (Fernandez et al., 2006).”
BackgroundVinardo: A Scoring Function Based on Autodock Vina Improves Scoring, Docking, and Virtual Screening
Quiroga, R., & Villarreal, M. A. (2016) · PLoS ONE · doi:10.1371/journal.pone.0155183 · Cited by 405 · cited 3× · see citing works
“We used LeDock version 1.0 (Zhao and Huang, 2011), rDock version 2013.1 (Ruiz-Carmona et al., 2014), AutoDock Vina version 1.1.2 (Trott and Olson, 2010), and smina version ‘‘Oct 15 2019’’ (Koes et al., 2013), the last of which implemented the DK and Vinardo (Quiroga and Villarreal, 2016) scoring functions in addition…”
BackgroundMethodInfluentialExploring genetic interaction manifolds constructed from rich single-cell phenotypes
Norman, T. M., Horlbeck, M. A., Replogle, J. M., Ge, A. Y., Xu, A., Jost, M., Gilbert, L. A., & Weissman, J. S. (2019) · Science · doi:10.1126/science.aax4438 · Cited by 392 · cited 3× · see citing works
“Many of these conditions induced differentiation of K562, including strong differentiated signatures consistent with an erythroid state (Norman et al., 2019) (Figure 7B).”
BackgroundMethodInfluentialMicchelli, C. A., Xu, Y., & Zhang, H. (2006) · Syracuse University Libraries (Syracuse University) · Cited by 379 · see citing works
Oppermann, H., Levinson, A. D., VARMUS, H. E., Levintow, L., & Bishop, J. M. (1979) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.76.4.1804 · Cited by 375 · cited 1× · see citing works
“…kinase involved in Toll-like receptor signaling (Kawagoe et al., 2007); human c-SRC, a tyrosine kinase and canonical proto-oncogene (Oppermann et al., 1979); human p110 d , a lipid kinase and leukocytic immune regulator (Vanhaesebroeck et al., 1997); and Mtb PknB, a…”
BackgroundDrug–target affinity prediction using graph neural network and contact maps
Jiang, M., Li, Z., Zhang, S., Wang, S., Wang, X., Yuan, Q., & Wei, Z. (2020) · RSC Advances · doi:10.1039/d0ra02297g · Cited by 372 · cited 7× · see citing works
“We used DGraphDTA (Jiang et al., 2020) to predict compound-kinase Kds. DGraphDTA leverages a graph neural network based on the compound molecular structure and the protein residue contact map.”
BackgroundMethodInfluentialOptimisation of Bioluminescent Reporters for Use with Mycobacteria
Andreu, N., Zelmer, A., Fletcher, T. I., Elkington, P. T., Ward, T. H., Ripoll, J., Parish, T., Bancroft, G. J., Schaible, U. E., Robertson, B. D., et al. (2010) · PLoS ONE · doi:10.1371/journal.pone.0010777 · Cited by 355 · cited 1× · see citing works
“Macrophages were infected with luxMtb and the luminescence is measured as a proxy of bacterial growth (Andreu et al., 2010; Bielecka et al., 2017) (STAR Methods).”
MethodThe Role of Src in Solid Tumors
Wheeler, D. L., Iida, M., & Dunn, E. F. (2009) · The Oncologist · doi:10.1634/theoncologist.2009-0009 · Cited by 346 · cited 2× · see citing works
“These kinases have well-documented roles in cancer, immunological, or infectious disease (Ali et al., 2014; Lougheed et al., 2011; Wang et al., 2009; Wheeler et al., 2009).”
BackgroundZheng, X., Ding, H., Mamitsuka, H., & Zhu, S. (2013) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/2487575.2487670 · Cited by 335 · see citing works
Edward: A library for probabilistic modeling, inference, and criticismPreprint
Tran, D., Kucukelbir, A., Dieng, A. B., Rudolph, M. R., Liang, D., & Blei, D. (2016) · arXiv.org · doi:10.48550/arxiv.1610.09787 · Cited by 311 · cited 3× · see citing works
“Optimization was performed under a mean-field independence assumption with gradient descent-based variational inference (Neal, 2012; Tran et al., 2016).”
MethodInfluentialIdentifying the proteins to which small-molecule probes and drugs bind in cells
Ong, S., Schenone, M., Margolin, A. A., Li, X., Do, K., Doud, M. K., Mani, D. R., Kuai, L., Wang, X., Wood, J. L., et al. (2009) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0900191106 · Cited by 304 · cited 1× · see citing works
“36 nM), automating discoveries that previously had been made with massive-scale screens or expert biochemical reasoning (Ali et al., 2014; Ong et al., 2009).”
BackgroundAssessing the impact of generative AI on medicinal chemistry
Walters, W. P., & Murcko, M. (2020) · Nature Biotechnology · doi:10.1038/s41587-020-0418-2 · Cited by 262 · see citing works
Compact integration of multi-network topology for functional analysis of genes
Cho, H., Berger, B., & Peng, J. (2016) · Cell Systems · doi:10.1016/j.cels.2016.10.017 · Cited by 241 · see citing works
Self-citationRelational Learning via Collective Matrix Factorization
Singh, A. P., & Gordon, G. J. (2008) · doi:10.21236/ada486804 · Cited by 240 · see citing works
Discovery and development of new antibacterial drugs: learning from experience?
Jackson, N. C., Czaplewski, L. G., & Piddock, L. J. V. (2018) · Journal of Antimicrobial Chemotherapy · doi:10.1093/jac/dky019 · Cited by 239 · see citing works
Learning protein sequence embeddings using information from structure
Bepler, T., & Berger, B. (2019) · arXiv · doi:10.48550/arxiv.1902.08661 · Cited by 219 · cited 2× · see citing works
“21 M protein sequences) (Bepler and Berger, 2019) (STAR Methods).”
MethodSelf-citationBarondeau, D. P., Putnam, C. D., Kassmann, C. J., Tainer, J. A., & Getzoff, E. D. (2003) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2133463100 · Cited by 215 · cited 1× · see citing works
“A notable exception is high GP prioritization of mutations to T62 (Figure 6E), a buried residue essential to chromophore synthesis (Barondeau et al., 2003).”
BackgroundA Deep Neural Network for Predicting and Engineering Alternative Polyadenylation
Bogard, N., Linder, J., Rosenberg, A. B., & Seelig, G. (2019) · Cell · doi:10.1016/j.cell.2019.04.046 · Cited by 215 · cited 1× · see citing works
“Machine learning algorithms offer a way to translate existing data into actionable biological hypotheses (Bogard et al., 2019; King et al., 2004; LeCun et al., 2015; Stokes et al., 2020; Tarca et al., 2007; Yang et al., 2019).”
BackgroundNorinder, U., Carlsson, L., Boyer, S. K., & Eklund, M. (2014) · Journal of Chemical Information and Modeling · doi:10.1021/ci5001168 · Cited by 208 · see citing works
Machine learning-guided channelrhodopsin engineering enables minimally invasive optogenetics
Bedbrook, C. N., Yang, K., Robinson, J. E., Mackey, E. D. W., Gradinaru, V., & Arnold, F. H. (2019) · Nature Methods · doi:10.1038/s41592-019-0583-8 · Cited by 200 · cited 1× · see citing works
“To demonstrate generality, we applied our same learning paradigm to predict the brightness of fluorescent proteins based on protein sequence features (STAR Methods), potentially enabling an algorithm that can optimize, in silico, the fluorescence of an existing protein design (Bedbrook et al., 2019; Yang et al., 2019).”
MethodThe Ser/Thr Protein Kinase PknB Is Essential for Sustaining Mycobacterial Growth
Fernandez, P., Saint‐Joanis, B., Barilone, N., Jackson, M., Gicquel, B., Cole, S., & Alzari, P. (2006) · Journal of Bacteriology · doi:10.1128/jb.00963-06 · Cited by 198 · cited 3× · see citing works
“…et al., 2007); human c-SRC, a tyrosine kinase and canonical proto-oncogene (Oppermann et al., 1979); human p110 d , a lipid kinase and leukocytic immune regulator (Vanhaesebroeck et al., 1997); and Mtb PknB, a serine-threonine kinase essential to mycobacterial viability (Fernandez et al., 2006).”
BackgroundPredicting Drug–Target Interactions Using Probabilistic Matrix Factorization
Çobanoğlu, M. C., Liu, C., Hu, F., Oltvai, Z. N., & Bahar, İ. (2013) · Journal of Chemical Information and Modeling · doi:10.1021/ci400219z · Cited by 187 · cited 2× · see citing works
“…on information from 72 compounds to make predictions over a 10,833-compound library is a more imbalanced train and test split than other reported drug-target interaction prediction settings (Cobanoglu et al., 2013; Hie et al., 2018; Luo et al., 2017; O¨zt € urk et al., 2018; Zheng et al., 2013).”
BackgroundMethodKawagoe, T., Sato, S., Jung, A., Yamamoto, M., Matsui, K., Kato, H., Uematsu, S., Takeuchi, O., & Akira, S. (2007) · The Journal of Experimental Medicine · doi:10.1084/jem.20061523 · Cited by 187 · cited 1× · see citing works
“Since our in vitro binding assays are optimized to screen many compounds for a given kinase, we focused our validation efforts on a set of four diverse kinases: human interleukin-1 receptor-associated kinase 4 (IRAK4), a serine-threonine kinase involved in Toll-like receptor signaling (Kawagoe et al., 2007); human c-S…”
BackgroundExponential consensus ranking improves the outcome in docking and receptor ensemble docking
Palacio‐Rodríguez, K., Lans, I., Cavasotto, C. N., & Cossio, P. (2019) · Scientific Reports · doi:10.1038/s41598-019-41594-3 · Cited by 186 · cited 2× · see citing works
“Since consistency across disparate docking scoring functions corresponds to better prediction of true biochemical affinity (Palacio-Rodrı́guez et al., 2019), we used six scoring functions to compare generated designs selected with and without uncertainty (Koes et al., 2013; Trott and Olson, 8 Cell…”
MethodWhy is my classifier discriminatory?
Chen, I., Johansson, F. D., & Sontag, D. (2018) · arXiv · doi:10.48550/arxiv.1805.12002 · Cited by 185 · see citing works
Deng, Y. C., Bao, F., Dai, Q., Wu, L. F., & Altschuler, S. J. (2019) · Nature Methods · doi:10.1038/s41592-019-0353-7 · Cited by 173 · cited 1× · see citing works
“Imputation is an important component of many functional genomics analyses such as noise reduction (Deng et al., 2019; van Dijk et al., 2018), compressed sensing 0 2 4 6 8 10 12 14 Number of mutations 0 2 4 6 8 10 12”
BackgroundGeometric Sketching Compactly Summarizes the Single-Cell Transcriptomic Landscape
Hie, B., Cho, H., DeMeo, B., Bryson, B. D., & Berger, B. (2019) · Cell Systems · doi:10.1016/j.cels.2019.05.003 · Cited by 170 · see citing works
Self-citationTowards the generalized iterative synthesis of small molecules
Lehmann, J. W., Blair, D. J., & Burke, M. D. (2018) · Nature Reviews Chemistry · doi:10.1038/s41570-018-0115 · Cited by 151 · cited 1× · see citing works
“Sample efficiency is especially critical in domains where new data collection is limited or slow (for example, synthesizing and testing customized small-molecule drugs (Lehmann et al., 2018)).”
BackgroundIRAK-4 Inhibitors for Inflammation
Wang, Z., Wesche, H., Stevens, T., Walker, N. P. C., & Yeh, W. (2009) · Current Topics in Medicinal Chemistry · doi:10.2174/156802609789044407 · Cited by 149 · cited 2× · see citing works
“These kinases have well-documented roles in cancer, immunological, or infectious disease (Ali et al., 2014; Lougheed et al., 2011; Wang et al., 2009; Wheeler et al., 2009).”
BackgroundBacterial tyrosine kinases: evolution, biological function and structural insights
Grangeasse, C., Nessler, S., & Mijakovic̀, I. (2012) · Philosophical Transactions of the Royal Society B Biological Sciences · doi:10.1098/rstb.2011.0424 · Cited by 147 · cited 1× · see citing works
“Bacterial kinases are less well studied than human (or mammalian) kinases (Grangeasse et al., 2012; Jackson et al., 2018) but are nonetheless important therapeutic targets ll OPEN ACCESS Article”
BackgroundDesign and preparation of 2-benzamido-pyrimidines as inhibitors of IKK
Waelchli, R., Bollbuck, B., Bruns, C., Buhl, T., Eder, J., Feifel, R., Hersperger, R., Janser, P., Révész, L., Zerwes, H., et al. (2005) · Bioorganic & Medicinal Chemistry Letters · doi:10.1016/j.bmcl.2005.09.035 · Cited by 144 · cited 9× · see citing works
“Although the GP uncertainty scores increased by as much as a factor of 2 from the first round (Figure S5), indicating hypotheses that explore riskier, more distal regions of the compound landscape,we still found that one of the GP-acquired compounds, IKK-16 (Waelchli et al., 2006), binds PknB with a Kd of 22 nM, the se…”
BackgroundInfluentialLiao, A. T., Chien, M. B., Shenoy, N. R., Mendel, D. B., McMahon, G., Cherrington, J. M., & London, C. A. (2002) · Blood · doi:10.1182/blood-2001-12-0350 · Cited by 124 · cited 5× · see citing works
“Interestingly, the latter two of these interactions do not seem to have existing experimental support; WS3 was developed as an inducer of pancreatic beta-cell proliferation (Shen et al., 2013) and SU11652 was developed for human receptor tyrosine kinase inhibition (Liao et al., 2002).”
BackgroundInfluentialWehenkel, A., Fernandez, P., Bellinzoni, M., Catherinot, V., Barilone, N., Labesse, G., Jackson, M., & Alzari, P. (2006) · FEBS Letters · doi:10.1016/j.febslet.2006.04.046 · Cited by 121 · cited 2× · see citing works
“For the compounds with strong binding affinity, visualizing the binding poses suggested by the docking algorithm showed concordance with a known crystal-lography-determined small-molecule pose (Wehenkel et al., 2006) (PDB: 2FUM) (Figure 5D).”
MethodPooled ChIP-Seq Links Variation in Transcription Factor Binding to Complex Disease Risk
Tehranchi, A. K., Myrthil, M., Martin, T., Hie, B. L., Golan, D., & Fraser, H. B. (2016) · Cell · doi:10.1016/j.cell.2016.03.041 · Cited by 120 · cited 1× · see citing works
“As high-throughput assays continue to transform biology (ICGC/ TCGA Pan-Cancer Analysis of Whole Genomes Consortium, 2020; Davis et al., 2011; Hie et al., 2020; Rood et al., 2019; Sarkisyan et al., 2016; Tehranchi et al., 2016), the ultimate goal of these studies remains the same—to generate hypotheses that elucidate…”
BackgroundSelf-citationHydrogen Bonding Penalty upon Ligand Binding
Zhao, H., & Huang, D. (2011) · PLoS ONE · doi:10.1371/journal.pone.0019923 · Cited by 117 · see citing works
Toward a Common Coordinate Framework for the Human Body
Rood, J., Stuart, T., Ghazanfar, S., Biancalani, T., Fisher, E., Butler, A., Hupalowska, A., Gaffney, L., Mauck, W. M., Eraslan, G., et al. (2019) · Cell · doi:10.1016/j.cell.2019.11.019 · Cited by 115 · cited 2× · see citing works
“As high-throughput assays continue to transform biology (ICGC/ TCGA Pan-Cancer Analysis of Whole Genomes Consortium, 2020; Davis et al., 2011; Hie et al., 2020; Rood et al., 2019; Sarkisyan et al., 2016; Tehranchi et al., 2016), the ultimate goal of these studies remains the same—to generate hypotheses that elucidate…”
BackgroundEfficient Generation of Transcriptomic Profiles by Random Composite Measurements
Cleary, B., Cong, L., Cheung, A., Lander, E. S., & Regev, A. (2017) · Cell · doi:10.1016/j.cell.2017.10.023 · Cited by 114 · cited 1× · see citing works
“…training a predictive model on the expression values of a set of genes (a feature set) to infer the expression of a different set of genes (a target set) (Figure 7A (Cleary et al., 2017), gene regulatory inference (Mueller et al., 2017), and cross-modality transfer learning (Zhou et al., 2020).”
BackgroundSmall-Molecule Inducer of β Cell Proliferation Identified by High-Throughput Screening
Shen, W., Tremblay, M. S., Deshmukh, V. A., Wang, W., Filippi, C. M., Harb, G., Zhang, Y.-Q., Kamireddy, A., Baaten, J., Jin, Q., et al. (2013) · Journal of the American Chemical Society · doi:10.1021/ja309304m · Cited by 97 · cited 2× · see citing works
“Interestingly, the latter two of these interactions do not seem to have existing experimental support; WS3 was developed as an inducer of pancreatic beta-cell proliferation (Shen et al., 2013) and SU11652 was developed for human receptor tyrosine kinase inhibition (Liao et al., 2002).”
BackgroundA Visual Exploration of Gaussian Processes
Görtler, J., Kehlbeck, R., & Deußen, O. (2019) · Distill · doi:10.23915/distill.00017 · Cited by 95 · cited 2× · see citing works
“Our first uncertainty model fits a GP regressor (Görtler et al., 2019; Rasmussen and Williams, 2005) to the training set.”
BackgroundVisualizing Proton Antenna in a High-Resolution Green Fluorescent Protein Structure
Shinobu, A., Palm, G. J., Schierbeek, A. J., & Agmon, N. (2010) · Journal of the American Chemical Society · doi:10.1021/ja1010652 · Cited by 92 · cited 1× · see citing works
“Mutations involved in the top hundred acquired sequences by the GP were located on an X-ray crystallography-determined avGFP structure (PDB: 2WUR) (Shinobu et al., 2010).”
BackgroundComputational Methods for Single-Cell RNA Sequencing
Hie, B., Peters, J. M., Nyquist, S. K., Shalek, A. K., Berger, B., & Bryson, B. D. (2020) · Annual Review of Biomedical Data Science · doi:10.1146/annurev-biodatasci-012220-100601 · Cited by 87 · cited 1× · see citing works
“As high-throughput assays continue to transform biology (ICGC/ TCGA Pan-Cancer Analysis of Whole Genomes Consortium, 2020; Davis et al., 2011; Hie et al., 2020; Rood et al., 2019; Sarkisyan et al., 2016; Tehranchi et al., 2016), the ultimate goal of these studies remains the same—to generate hypotheses that elucidate…”
BackgroundSelf-citationEffective inhibitors of the essential kinase PknB and their potential as anti-mycobacterial agents
Lougheed, K. E. A., Osborne, S., Saxty, B., Whalley, D. P., Chapman, T., Bouloc, N., Chugh, J., Nott, T., Patel, D., Spivey, V., et al. (2011) · Tuberculosis · doi:10.1016/j.tube.2011.03.005 · Cited by 85 · cited 3× · see citing works
“These kinases have well-documented roles in cancer, immunological, or infectious disease (Ali et al., 2014; Lougheed et al., 2011; Wang et al., 2009; Wheeler et al., 2009).”
BackgroundMethodSurface protein imputation from single cell transcriptomes by deep neural networks
Zhou, Z., Ye, C., Wang, J., & Zhang, N. R. (2020) · Nature Communications · doi:10.1038/s41467-020-14391-0 · Cited by 83 · see citing works
Realizing private and practical pharmacological collaboration
Hie, B., Cho, H., & Berger, B. (2018) · Science · doi:10.1126/science.aat4807 · Cited by 76 · cited 2× · see citing works
“On the same neural pretrained features, we test an MLP, also known as a densely connected neural network (Hie et al., 2018; Özt€ urk et al., 2018) and collective matrix factorization (CMF), a model often used to recommend shopping items for potential buyers that can also recommend drugs for potential targets (Cobanogl…”
MethodSelf-citationMycobacterium tuberculosis Ser/Thr Protein Kinase B Mediates an Oxygen-Dependent Replication Switch
Ortega, C., Liao, R., Anderson, L. N., Rustad, T., Ollodart, A. R., Wright, A. T., Sherman, D. R., & Grundner, C. (2014) · PLoS Biology · doi:10.1371/journal.pbio.1001746 · Cited by 74 · cited 1× · see citing works
“In examining the literature for prior work on compounds targeting PknB, we identified support for K252a as an inhibitor of PknB kinase activity and Mtb growth (Fernandez et al., 2006; Ortega et al., 2014).”
BackgroundA review of active learning approaches to experimental design for uncovering biological networks
Sverchkov, Y., & Craven, M. W. (2017) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1005466 · Cited by 73 · cited 3× · see citing works
“This is the iterative cycle of computation and experimentation at the heart of active learning (Eisenstein, 2020; Sverchkov and Craven, 2017), for which we provide a proof-of-concept example in this study.”
BackgroundMethodZeng, H., & Gifford, D. K. (2019) · Cell Systems · doi:10.1016/j.cels.2019.05.004 · Cited by 69 · cited 2× · see citing works
“…et al., 2017; Neal, 2012), in the biological setting, many machine learning studies do not consider uncertainty or are limited to specific tasks or in-silico validations (Cortes-Ciriano et al., 2015; Ewing et al., 1998; Norinder et al., 2014; Sverchkov and Craven, 2017; Zeng and Gifford, 2019).”
BackgroundBielecka, M. K., Tezera, L. B., Zmijan, R., Drobniewski, F. A., Zhang, X., Jayasinghe, S. N., & Elkington, P. T. (2017) · mBio · doi:10.1128/mbio.02073-16 · Cited by 65 · cited 1× · see citing works
“Macrophages were infected with luxMtb and the luminescence is measured as a proxy of bacterial growth (Andreu et al., 2010; Bielecka et al., 2017) (STAR Methods).”
MethodSample Efficient Reinforcement Learning with Gaussian Processes
Grande, R. C., Walsh, T. J., & How, J. P. (2014) · Cited by 41 · see citing works
Cortés‐Ciriano, I., Bender, A., & Malliavin, T. E. (2015) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.5b00101 · Cited by 37 · cited 1× · see citing works
“…et al., 2017; Neal, 2012), in the biological setting, many machine learning studies do not consider uncertainty or are limited to specific tasks or in-silico validations (Cortes-Ciriano et al., 2015; Ewing et al., 1998; Norinder et al., 2014; Sverchkov and Craven, 2017; Zeng and Gifford, 2019).”
BackgroundQiu, X., Meyerson, E., & Miikkulainen, R. P. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.00588 · Cited by 37 · cited 2× · see citing works
“Our second method first fits a multilayer perceptron (MLP), followed by fitting a GP to the residuals of the MLP predictions (MLP + GP) (Qiu et al., 2020).”
BackgroundMethodActive machine learning helps drug hunters tackle biology
Eisenstein, M. S. (2020) · Nature Biotechnology · doi:10.1038/s41587-020-0521-4 · Cited by 34 · cited 3× · see citing works
“This is the iterative cycle of computation and experimentation at the heart of active learning (Eisenstein, 2020; Sverchkov and Craven, 2017), for which we provide a proof-of-concept example in this study.”
MethodCold-start recommendations in Collective Matrix FactorizationPreprint
Cortes, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.00366 · Cited by 21 · cited 1× · see citing works
“The CMF objective was fit using the limited-memory Broyden–Fletcher–Goldfarb– Shanno algorithm (L-BFGS) via the cmfrec Python package version 0.5.3 (Cortes, 2018) (https://cmfrec.readthedocs.io/en/latest/).”
MethodInfluentialLearning Optimal InterventionsPreprint
Mueller, J. W., Reshef, D. N., Du, G., & Jaakkola, T. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.05027 · Cited by 3 · cited 5× · see citing works
“As data points become more distal to the training set, the GP uncertainty also grows to approach the prior uncertainty, analogous to human uncertainty increasing on examples that deviate from existing knowledge (Figure 1) (Bernardo and Smith, 2009; Grande et al., 2014; Mueller et al., 2017; Rasmussen and Williams, 200…”
BackgroundInfluential
Flagged first, then by citations
6 of the 125 references couldn’t be matched to a known paper.
Research works printed but not linked: 1
- Bernardo, J.M., and Smith, A.F.M. (2009). Bayesian Theory (John Wiley & Sons, Ltd).
Cited by
Sorted by citations
Learning the protein language: Evolution, structure, and function
Bepler, T., & Berger, B. (2021) · Cell Systems · doi:10.1016/j.cels.2021.05.017 · Cited by 483 · see citing works
Self-citationEfficient evolution of human antibodies from general protein language models
Hie, B., Shanker, V. R., Xu, D., Bruun, T. U. J., Weidenbacher, P. A. B., Tang, S., Wu, W. C. H., Pak, J. E., & Kim, P. S. (2023) · Nature Biotechnology · doi:10.1038/s41587-023-01763-2 · Cited by 444 · see citing works
Self-citationEfficient Exploration of Chemical Space with Docking and Deep Learning
Ying, Y., Yao, K., Repasky, M. P., Leswing, K., Abel, R., Shoichet, B. K., & Jerome, S. V. (2021) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.1c00810 · Cited by 454 · see citing works
How to build the virtual cell with artificial intelligence: Priorities and opportunities
Bunne, C., Roohani, Y., Rosen, Y., Gupta, A., Zhang, X., Roed, M., Alexandrov, T., AlQuraishi, M., Brennan, P., Burkhardt, D. B., et al. (2024) · Cell · doi:10.1016/j.cell.2024.11.015 · Cited by 393 · see citing works
Generative models for molecular discovery: Recent advances and challenges
Bilodeau, C., Jin, W., Jaakkola, T., Barzilay, R., & Jensen, K. F. (2022) · Wiley Interdisciplinary Reviews Computational Molecular Science · doi:10.1002/wcms.1608 · Cited by 334 · see citing works
Opportunities and Challenges for Machine Learning-Assisted Enzyme Engineering
Yang, J., Li, F.-Z., & Arnold, F. H. (2024) · ACS Central Science · doi:10.1021/acscentsci.3c01275 · Cited by 257 · see citing works
“31,132,208,209 Broadly, active learning is an iterative cycle that alternates between wet-lab experiments to synthesize/screen enzymes and computational modeling to propose the next set of enzymes to test, typically guided by uncertainty quantification.”
BackgroundEvidential Deep Learning for Guided Molecular Property Prediction and Discovery
Soleimany, A. P., Amini, A., Goldman, S., Rus, D. L., Bhatia, S. N., & Coley, C. W. (2021) · ACS Central Science · doi:10.1021/acscentsci.1c00546 · Cited by 231 · see citing works
“In contrast, NNs can be trained to predict the parameters of the underlying probability distribution and obtain closed-form solutions of aleatoric (data) uncertainty, without sampling [12, 2, 7, 8].”
Sledzieski, S., Singh, R., Cowen, L., & Berger, B. (2021) · Cell Systems · doi:10.1016/j.cels.2021.08.010 · Cited by 208 · see citing works
Self-citationInformed training set design enables efficient machine learning-assisted directed protein evolution
Wittmann, B. J., Yue, Y., & Arnold, F. H. (2021) · Cell Systems · doi:10.1016/j.cels.2021.07.008 · Cited by 208 · see citing works
Machine Learning-Guided Protein Engineering
Kouba, P., Kohout, P., Haddadi, F., Bushuiev, A., Samusevich, R., Sedlář, J., Damborský, J., Pluskal, T., Šivic, J., & Mazurenko, S. (2023) · ACS Catalysis · doi:10.1021/acscatal.3c02743 · Cited by 214 · see citing works
“187 A recent advance in the area of active learning is the development of GFlowNets, 188 networks designed to suggest diverse and accurate candidates in a machine − expert loop to accelerate scientific discovery.”
Singh, R., Sledzieski, S., Bryson, B. D., Cowen, L., & Berger, B. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2220778120 · Cited by 207 · see citing works
Self-citationAdvances in machine learning for directed evolution
Wittmann, B. J., Johnston, K. E., Wu, Z., & Arnold, F. H. (2021) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2021.01.008 · Cited by 165 · see citing works
Deep generative models for peptide design
Wan, F., Kontogiorgos-Heintz, D., & de la Fuente‐Núñez, C. (2022) · Digital Discovery · doi:10.1039/d1dd00024a · Cited by 150 · see citing works
“109 In return, the experimental results of selected samples can be used as feedback to re ne generative and property predictor models by e ffi ciently expanding training data space and reducing model uncertainty.”
Protein sequence design with deep generative models
Wu, Z., Johnston, K. E., Arnold, F. H., & Yang, K. (2021) · Current Opinion in Chemical Biology · doi:10.1016/j.cbpa.2021.04.004 · Cited by 141 · see citing works
“Additionally, while studies incorporating knowledge of protein structure are becoming increasingly powerful [9, 10, 11, 12, 13], they are beyond the scope of this review, and we focus on deep generative models of protein sequence.”
BackgroundActive learning-assisted directed evolution
Yang, J., Lal, R., Bowden, J. C., Astudillo, R., Hameedi, M. A., Kaur, S., Hill, M., Yue, Y., & Arnold, F. H. (2025) · Nature Communications · doi:10.1038/s41467-025-55987-8 · Cited by 144 · see citing works
“15,26 Overall, classical notions of uncertainty quantification seem to play a more limited role than expected in these real-world applications.”
Defining and Exploring Chemical Spaces
Coley, C. W. (2020) · Trends in Chemistry · doi:10.1016/j.trechm.2020.11.004 · Cited by 126 · see citing works
Machine learning modeling of family wide enzyme-substrate specificity screens
Goldman, S., Das, R., Yang, K. K., & Coley, C. W. (2022) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1009853 · Cited by 120 · see citing works
CatPred: a comprehensive framework for deep learning in vitro enzyme kinetic parameters
Boorla, V. S., & Maranas, C. D. (2025) · Nature Communications · doi:10.1038/s41467-025-57215-9 · Cited by 114 · see citing works
“For example, Hie et al. 42 used a Gaussian-process based UQ to discover several kinase inhibitors with nanomolar binding af fi - nities.”
Adaptive machine learning for protein engineering
Hie, B., & Yang, K. (2021) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2021.11.002 · Cited by 116 · see citing works
Self-citationEvaluating pointwise reliability of machine learning prediction
Nicora, G., Rios, M., Abu–Hanna, A., & Bellazzi, R. (2022) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2022.103996 · Cited by 98 · see citing works
Genome-wide mapping of somatic mutation rates uncovers drivers of cancer
Sherman, M. A., Yaari, A. U., Priebe, O., Dietlein, F., Loh, P.-R., & Berger, B. (2022) · Nature Biotechnology · doi:10.1038/s41587-022-01353-8 · Cited by 80 · see citing works
Self-citationDing, K., Chin, M., Zhao, Y., Huang, W., Mai, B. K., Wang, H., Liu, P., Yang, Y., & Luo, Y. (2024) · Nature Communications · doi:10.1038/s41467-024-50698-y · Cited by 86 · see citing works
The future of antibiotics begins with discovering new combinations
Zhu, M., Tse, M. W., Weller, J., Chen, J., & Blainey, P. C. (2021) · Annals of the New York Academy of Sciences · doi:10.1111/nyas.14649 · Cited by 86 · see citing works
Obrezanova, O., Martinsson, A., Whitehead, T., Mahmoud, S. Y., Bender, A., Miljković, F., Grabowski, P., Irwin, B. W. J., Oprisiu, I., Conduit, G., et al. (2022) · Molecular Pharmaceutics · doi:10.1021/acs.molpharmaceut.2c00027 · Cited by 74 · see citing works
Larkins‐Ford, J., Greenstein, T., Van, N., Degefu, Y. N., Olson, M. C., Sokolov, A., & Aldridge, B. B. (2021) · Cell Systems · doi:10.1016/j.cels.2021.08.004 · Cited by 76 · see citing works
“We expect this pipeline to improve as the component drug set is expanded and diversified via new drug discovery methods (Hie et al., 2020; Johnson et al., 2019; Rock, 2019; Rock et al., 2017; VanderVen et al., 2015; Bryk et al., 2008; Gold and Nathan, 2017; Gold et al., 2015) and as more animal studies provide outcome…”
Artificial intelligence for compound pharmacokinetics prediction
Obrezanova, O. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102546 · Cited by 74 · see citing works
Artificial intelligence for science in quantum, atomistic, and continuum systems
Zhang, X., Wang, L., Helwig, J., Luo, Y., Fu, C., Xie, Y., Liu, M., Lin, Y., Zhao, X., Yan, K., et al. (2025) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000115 · Cited by 123 · see citing works
Kumar, S. A., Kumar, T. D. A., Beeraka, N. M., Pujar, G. V., Singh, M., Akshatha, H. S., & Bhagyalalitha, M. (2021) · Future Medicinal Chemistry · doi:10.4155/fmc-2021-0243 · Cited by 71 · see citing works
Self-play reinforcement learning guides protein engineering
Wang, Y., Tang, H., Huang, L., Pan, L., Yang, L., Yang, H., Mu, F., & Yang, M. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00691-9 · Cited by 69 · see citing works
“Bayesian optimization (BO) is an alternative to greedy acquisition 22 , which estimates model uncertainty and balances exploration and exploitation through upper confidence bound acquisition 23 .”
BackgroundAssigning confidence to molecular property prediction
Nigam, A., Pollice, R., Hurley, M. F. D., Hickman, R. J., Aldeghi, M., Yoshikawa, N., Chithrananda, S., Voelz, V. A., & Aspuru‐Guzik, A. (2021) · Expert Opinion on Drug Discovery · doi:10.1080/17460441.2021.1925247 · Cited by 68 · see citing works
Antibiotic discovery in the artificial intelligence era
Lluka, T., & Stokes, J. M. (2022) · Annals of the New York Academy of Sciences · doi:10.1111/nyas.14930 · Cited by 69 · see citing works
“50 As mentioned, given the complexity of chemical and biological systems, collecting large and high-quality datasets is essential tominimizepredictionerrorandmaximizegeneralizationinreal-world settings.”
BackgroundProtein engineering via Bayesian optimization-guided evolutionary algorithm and robotic experiments
Hu, R., Fu, L., Chen, Y., Chen, J., Qiao, Y., & Si, T. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac570 · Cited by 68 · see citing works
Calibrated geometric deep learning improves kinase–drug binding predictions
Luo, Y., Liu, Y., & Peng, J. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00751-0 · Cited by 67 · see citing works
“We first assessed KDBNet’s performance in predicting kinase–drug binding affinity using two public datasets of experimental measurements of kinase–compound binding affinity, Davis 27 and KIBA 28 , which were widely used to benchmark previous methods 10,12,15–17 .”
InfluentialGhislat, G., Rahman, T., & Ballester, P. J. (2021) · Current Opinion in Chemical Biology · doi:10.1016/j.cbpa.2021.04.009 · Cited by 66 · see citing works
Cluster learning-assisted directed evolution
Qiu, Y., Hu, J., & Wei, G. (2021) · Nature Computational Science · doi:10.1038/s43588-021-00168-y · Cited by 63 · see citing works
“Supervised models can learn relationships between sequences and fitness properties, and provide quantitative predictions on protein thermostability [7], protein folding energy [8, 9], protein solubility [10], protein-ligand binding affinity [11], and protein-protein binding affinity [12].”
BackgroundUncertainty quantification: Can we trust artificial intelligence in drug discovery?
Yu, J., Wang, D., & Zheng, M. (2022) · iScience · doi:10.1016/j.isci.2022.104814 · Cited by 59 · see citing works
“valid the effectiveness of this strategy based on the task of modeling compound-kinase interaction (Hie et al., 2020).”
BackgroundIn vitro continuous protein evolution empowered by machine learning and automation
Yu, T., Boob, A. G., Singh, N., Su, Y., & Zhao, H. (2023) · Cell Systems · doi:10.1016/j.cels.2023.04.006 · Cited by 57 · see citing works
Discovery of potent inhibitors of α-synuclein aggregation using structure-based iterative learning
Horne, R. I., Andrzejewska, E. A., Alam, P., Brotzakis, Z. F., Srivastava, A., Aubert, A., Nowinska, M., Gregory, R. C., Staats, R., Possenti, A., et al. (2024) · Nature Chemical Biology · doi:10.1038/s41589-024-01580-x · Cited by 56 · see citing works
“While this use of a structurally related screening library constrains the model’s ability to generalize, the lack of diversity in terms of potent molecules in the training set also makes it unlikely for the model to perform well in chemical space divergent from this region.”
InfluentialBusk, J., Jørgensen, P. B., Bhowmik, A., Schmidt, M. N., Winther, O., & Vegge, T. (2021) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ac3eb3 · Cited by 59 · see citing works
“When quantifying uncertainty it is often useful to distinguish between epistemic and aleatoric uncertainty [17, 18].”
BackgroundCross-disciplinary perspectives on the potential for artificial intelligence across chemistry
Mroz, A. M., Basford, A. R., Hastedt, F., Jayasekera, I. S., Mosquera‐Lois, I., Sedgwick, R., Ballester, P. J., Bocarsly, J. D., del Rio‐Chanona, E. A., Evans, M. L., et al. (2025) · Chemical Society Reviews · doi:10.1039/d5cs00146c · Cited by 59 · see citing works
Neural network extrapolation to distant regions of the protein fitness landscape
Freschlin, C. R., Fahlberg, S. A., Heinzelman, P., & Romero, P. A. (2024) · Nature Communications · doi:10.1038/s41467-024-50712-3 · Cited by 50 · see citing works
“10 and 11).”
Advancing Targeted Protein Degradation via Multiomics Profiling and Artificial Intelligence
Duran-Frigola, M., Cigler, M., & Winter, G. E. (2023) · Journal of the American Chemical Society · doi:10.1021/jacs.2c11098 · Cited by 50 · see citing works
Active learning accelerates electrolyte solvent screening for anode-free lithium metal batteries
Ma, P., Kumar, R., Wang, K.-H., & Amanchukwu, C. V. (2025) · Nature Communications · doi:10.1038/s41467-025-63303-7 · Cited by 48 · see citing works
“Version 1:”
Ghislat, G., Hernández-Hernández, S., Piyawajanusorn, C., & Ballester, P. J. (2024) · Expert Opinion on Drug Discovery · doi:10.1080/17460441.2024.2403639 · Cited by 52 · see citing works
“For example, Gaussian Process (GP)-based uncertainty prediction was employed to anticipate likely and potent small-molecule inhibitors of several molecular and phenotypic targets [79].”
Machine learning applications for therapeutic tasks with genomics data
Huang, K., Xiao, C., Glass, L. M., Critchlow, C. W., Gibson, G. C., & Sun, J. (2021) · Patterns · doi:10.1016/j.patter.2021.100328 · Cited by 46 · see citing works
“Recently, Hie et al.102 used Gaussian processes to generate uncertainty scores of compound bioactivity, protein fluorescence, and single-cell transcriptomic imputation, which were shown to guide the experimental and validation loop.”
BackgroundMethodInfluentialWang, D., Yu, J., Chen, L., Li, X., Jiang, H., Chen, K., Zheng, M., & Luo, X. (2021) · Journal of Cheminformatics · doi:10.1186/s13321-021-00551-x · Cited by 46 · see citing works
Calibration and generalizability of probabilistic models on low-data chemical datasets with DIONYSUS
Tom, G., Hickman, R. J., Zinzuwadia, A. N., Mohajeri, A., Sánchez-Lengeling, B., & Aspuru‐Guzik, A. (2023) · Digital Discovery · doi:10.1039/d2dd00146b · Cited by 47 · see citing works
Impact of Artificial Intelligence on Compound Discovery, Design, and Synthesis
Miljković, F. N., Rodríguez-Pérez, R., & Bajorath, J. (2021) · ACS Omega · doi:10.1021/acsomega.1c05512 · Cited by 39 · see citing works
Narayanan, H., Hinckley, J. A., Barry, R. M., Dang, B., Wolffe, L. A., Atari, A., Tseng, Y., & Love, J. C. (2025) · Nature Communications · doi:10.1038/s41467-025-61113-5 · Cited by 36 · see citing works
PyPEF—An Integrated Framework for Data-Driven Protein Engineering
Siedhoff, N. E., Illig, A.-M., Schwaneberg, U., & Davari, M. D. (2021) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.1c00099 · Cited by 35 · see citing works
Active Learning for Drug Design: A Case Study on the Plasma Exposure of Orally Administered Drugs
Ding, X., Cui, R., Yu, J., Liu, T., Zhu, T., Wang, D., Chang, J., Fan, Z., Liu, X., Chen, K., et al. (2021) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.1c01683 · Cited by 36 · see citing works
Meta learning addresses noisy and under-labeled data in machine learning-guided antibody engineering
Minot, M., & Reddy, S. T. (2024) · Cell Systems · doi:10.1016/j.cels.2023.12.003 · Cited by 35 · see citing works
The long road to calibrated prediction uncertainty in computational chemistry
Pernot, P. (2022) · The Journal of Chemical Physics · doi:10.1063/5.0084302 · Cited by 35 · see citing works
Effective engineering of a ketoreductase for the biocatalytic synthesis of an ipatasertib precursor
Malca, S. H., Duss, N., Meierhofer, J., Patsch, D., Niklaus, M., Reiter, S., Hanlon, S. P., Wetzl, D., Kuhn, B., Iding, H., et al. (2024) · Communications Chemistry · doi:10.1038/s42004-024-01130-5 · Cited by 32 · see citing works
Deep Learning Methods for Small Molecule Drug Discovery: A Survey
Hu, W., Liu, Y., Chen, X., Chai, W., Chen, H., Wang, H., & Wang, G. (2023) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2023.3251977 · Cited by 30 · see citing works
“In addition, the algorithm using uncertainty quantization is robust when exploring the domain outside the distribution of training data, which is important for the model to explore in the huge and diverse chemical space [179].”
MethodWang, D. Y. Q., Hensman, J. J., Kutkaite, G., Toh, T. S., Galhoz, A., GDSC Screening Team, Lightfoot, H. W., Yang, W., Soleimani, M., Barthorpe, S., et al. (2020) · eLife · doi:10.7554/elife.60352 · Cited by 28 · see citing works
Benchmarking compound activity prediction for real-world drug discovery applications
(2024) · doi:10.1038/s42004-024-01204-4 · Cited by 34 · see citing works
“It is also important to further investigate whether it is possible to estimatesample-levelcon fi denceofthepredictionresults,whichisrelatedto the concept of uncertainty estimation 62 .”
Auerbach, B. J., Hu, J., Reilly, M. P., & Li, M. (2021) · Genome Research · doi:10.1101/gr.275430.121 · Cited by 25 · see citing works
“As other papers have thoroughly reviewed these aspects (Bacher and Kendziorski 2016; Hie et al. 2020; Hou et al. 2020; Lytal et al. 2020; Wu and Zhang 2020; Zhang and Zhang 2020; Ahlmann-Eltze and Huber 2021; Melsted et al. 2021; Slovin et al. 2021), wewill focus our discussion on the downstream…”
BackgroundBenchmarking uncertainty quantification for protein engineering
Greenman, K. P., Amini, A. P., & Yang, K. (2025) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012639 · Cited by 26 · see citing works
“While some protein engineering work has leveraged uncertainty estimates, these studies have been limited to single UQ methods such as ensembles (Mariet et al., 2020) or Gaussian processes (GPs) (Hie et al., 2020).”
BackgroundAhmad, B., Ouahada, K., & Hamam, H. (2026) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2025.12.033 · Cited by 29 · see citing works
“…curves to quan tify whether highlighted atoms / residues truly drive predictions; (ii) perturbation-based causal probes to verify attribution consistency; and (iii) uncertainty quantification to calibrate decision thresholds under class imbalance and distribution shift [153,154].”
Kim, Y., Wang, K., Lock, R. I., Nash, T. R., Fleischer, S., Wang, B. Z., Fine, B. M., & Vunjak‐Novakovic, G. (2024) · IEEE Open Journal of Engineering in Medicine and Biology · doi:10.1109/ojemb.2024.3377461 · Cited by 24 · see citing works
“There is also a need to facilitate ML applications to obtain more meaningful information from complex functional data, as evidenced by its utilization in drug discovery, medical di-agnostics, protein design, and sequencing analysis [12], [13], [14], [15], [16], [17].”
BackgroundLi, P., Hua, L., Ma, Z., Hu, W., Liu, Y., & Zhu, J. (2024) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.4c01139 · Cited by 25 · see citing works
Multilingual translation for zero-shot biomedical classification using BioTranslator
Xu, H., Woicik, A., Poon, H., Altman, R. B., & Wang, S. (2023) · Nature Communications · doi:10.1038/s41467-023-36476-2 · Cited by 24 · see citing works
“7, 8). g The generated text from BioTranslator given the proteins of phosphatidylglycerol metabolic process (GO:0046471) and the proteins of positive regulation of microtubule polymerization (GO:0031116).”
MethodArtificial Intelligence in Interdisciplinary Life Science and Drug Discovery Research
Bajorath, J. (2022) · Future Science OA · doi:10.2144/fsoa-2022-0010 · Cited by 23 · see citing works
“Although there are ML methods that provide inherent prediction uncertainties such as Gaussian process modeling [20] , most ML / DL methods including DNNs produce numerical end points without uncertainty estimates.”
MethodMachine Learning for Protein Engineering
Johnston, K. E., Fannjiang, C., Wittmann, B. J., Hie, B., Yang, K., & Wu, Z. (2023) · Challenges and advances in computational chemistry and physics · doi:10.1007/978-3-031-37196-7_9 · Cited by 23 · see citing works
Self-citationDeng, J., Gu, M., Zhang, P., Dong, M., Liu, T., Zhang, Y., & Liu, M. (2024) · Nature Machine Intelligence · doi:10.1038/s42256-024-00940-5 · Cited by 21 · see citing works
“Both goals ultimately return to reducing epistemic uncertainty 46,47 , which often comes from a lack of knowledge.”
Leveraging Immunopeptidomics To Study and Combat Infectious Disease
Leddy, O. K., White, F. M., & Bryson, B. D. (2021) · mSystems · doi:10.1128/msystems.00310-21 · Cited by 20 · see citing works
Self-citationLearning Strategies in Protein Directed Evolution
Cadet, X. F., Gelly, J., van Noord, A., Cadet, F., & Acevedo‐Rocha, C. G. (2022) · Methods in molecular biology · doi:10.1007/978-1-0716-2152-3_15 · Cited by 21 · see citing works
Protease Activity Analysis: A Toolkit for Analyzing Enzyme Activity Data
Soleimany, A. P., Martin-Alonso, C., Anahtar, M., Wang, C. S., & Bhatia, S. N. (2022) · ACS Omega · doi:10.1021/acsomega.2c01559 · Cited by 20 · see citing works
Gruich, C., Madhavan, V., Wang, Z., & Goldsmith, B. R. (2023) · Machine Learning Science and Technology · doi:10.1088/2632-2153/accace · Cited by 18 · see citing works
Active learning of enhancers and silencers in the developing neural retina
Friedman, R. Z., Ramu, A., Lichtarge, S., Wu, Y., Tripp, L., Lyon, D., Myers, C. A., Granas, D. M., Gause, M., Corbo, J. C., et al. (2025) · Cell Systems · doi:10.1016/j.cels.2024.12.004 · Cited by 17 · see citing works
Artificial intelligence-guided strategies for next-generation biological sequence design
Zhang, P., Wei, L., Li, J., & Wang, X. (2024) · National Science Review · doi:10.1093/nsr/nwae343 · Cited by 20 · see citing works
RECOVER identifies synergistic drug combinations in vitro through sequential model optimization
Bertin, P., Rector-Brooks, J., Sharma, D. K., Gaudelet, T., Anighoro, A., Groß, T., Martínez‐Peña, F., Tang, E. L., Suraj, M. S., Regep, C., et al. (2023) · Cell Reports Methods · doi:10.1016/j.crmeth.2023.100599 · Cited by 16 · see citing works
Hu, J., Liu, Y., Zeng, X., Zou, Q., Su, R., & Wei, L. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3553217 · Cited by 15 · see citing works
Fan, W., Zeng, L., & Wang, T. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.106180 · Cited by 15 · see citing works
Horne, R. I., Wilson-Godber, J., Díaz, A. G., Brotzakis, Z. F., Seal, S., Gregory, R. C., Possenti, A., Chia, S., & Vendruscolo, M. (2024) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c01777 · Cited by 15 · see citing works
Kim, J. Y., Rosenberger, M. G., Rutledge, N. S., & Esser-Kahn, A. P. (2023) · Pharmaceutics · doi:10.3390/pharmaceutics15061687 · Cited by 15 · see citing works
Ji, G., Li, S., Ye, L., & Guan, J. (2021) · Biomedicines · doi:10.3390/biomedicines9040410 · Cited by 13 · see citing works
Zhan, H., Zhu, X. Y., & Hu, J. (2022) · Applied Soft Computing · doi:10.1016/j.asoc.2022.109858 · Cited by 13 · see citing works
Overfit deep neural network for predicting drug-target interactions
Xiaolin, X., Liu, X., He, G., Hongwei, L., Jinkuo, G., Bian, X., Zhen, T., Ma, X., Li, Y., Xue, N., et al. (2023) · iScience · doi:10.1016/j.isci.2023.107646 · Cited by 13 · see citing works
A systematic analysis of regression models for protein engineering
Richard, M., Kæstel‐Hansen, J., Groth, P. M., Bartels, S., Salomon, J., Tian, P., Hatzakis, N. S., & Boomsma, W. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012061 · Cited by 12 · see citing works
Efforts to Minimise the Bacterial Genome as a Free-Living Growing System
Aida, H., & Ying, B. (2023) · Biology · doi:10.3390/biology12091170 · Cited by 10 · see citing works
Biophysics-guided uncertainty-aware deep learning uncovers high-affinity plastic-binding peptides
Alshehri, A. S., Bergman, M. T., You, F., & Hall, C. K. (2025) · Digital Discovery · doi:10.1039/d4dd00219a · Cited by 11 · see citing works
Reducing overconfident errors in molecular property classification using Posterior Network
Fan, Z., Yu, J., Zhang, X., Chen, Y., Sun, S., Zhang, Y., Chen, M., Xiao, F., Wu, W., Li, X., et al. (2024) · Patterns · doi:10.1016/j.patter.2024.100991 · Cited by 10 · see citing works
Zhang, P.-C., Du, Q.-X., Wang, Y., Wei, L., & Wang, X.-W. (2025) · Nature Communications · doi:10.1038/s41467-025-57031-1 · Cited by 10 · see citing works
“Therefore, using randomly sampled 5 ’ RSs from other species to represent the functional subspace of the target species is ineffective 12 .”
Predicting high-fitness viral protein variants with Bayesian active learning and biophysics
Huot, M., Wang, D., Liu, J., & Shakhnovich, E. I. (2025) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2503742122 · Cited by 10 · see citing works
Active learning-guided optimization of cell-free biosensors for lead testing in drinking water
Wang, B. M., Chiang, N., Ekas, H. M., Brown, D. M., Dildine, G., Lucci, T. J., Feng, S., Bly, V., Gaillard, J., Lucks, J. B., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-66964-6 · Cited by 12 · see citing works
“Machine learning (ML)-guided directed evolution has transformed the way we navigate vast protein sequence-function land-scapes, making exploration faster and less reliant on exhaustive experiments 31 – 39 .”
Parkinson, J., & Wang, W. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.3c00601 · Cited by 10 · see citing works
Wang, D., Wu, Z., Shen, C., Bao, L., Luo, H., Wang, Z., Yao, H., De, K., Luo, C., & Hou, T. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac592 · Cited by 9 · see citing works
Toward Guided Mutagenesis: Gaussian Process Regression Predicts MHC Class II Antigen Mutant Binding
Bell, D. R., & Chen, S. H. (2021) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.1c00458 · Cited by 9 · see citing works
Abbasi, A., & Nataraj, C. (2022) · IEEE Internet Computing · doi:10.1109/mic.2022.3170736 · Cited by 9 · see citing works
Wang, C., Anglès, F., & Balch, W. E. (2022) · Structure · doi:10.1016/j.str.2022.05.011 · Cited by 8 · see citing works
A View on Model Misspecification in Uncertainty Quantification
Kato, Y., Tax, D. M. J., & Loog, M. (2023) · Communications in computer and information science · doi:10.1007/978-3-031-39144-6_5 · Cited by 8 · see citing works
Wang, C., Elghobashi-Meinhardt, N., & Balch, W. (2022) · bioRxiv · doi:10.1101/2022.01.07.475295 · Cited by 6 · see citing works
“GPR also generates robust uncertainty for each prediction (51-54), 7 providing a rigorous tool to assess the probability of function of a residue in the context of other 8 evolving residues in response to the local environment (48).”
MethodLongitudinal profiling of low-abundance strains in microbiomes with ChronoStrain
Kim, Y., Worby, C. J., Acharya, S., van Dijk, L. R., Alfonsetti, D., Gromko, Z., Azimzadeh, P. N., Dodson, K., Gerber, G. K., Hultgren, S. J., et al. (2025) · Nature Microbiology · doi:10.1038/s41564-025-01983-z · Cited by 7 · see citing works
Self-citationBadkul, A., Xie, L., Zhang, S., & Xie, L. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-025-01151-2 · Cited by 7 · see citing works
“Several works 36,37 propose a combined GP and multilayer perceptron (MLP) approach for various biological tasks.”
Shil, R., Pattapulavar, V., Ramanujam, S., Subburaj, S., & Christopher, J. G. (2026) · ACS Omega · doi:10.1021/acsomega.5c08473 · Cited by 7 · see citing works
Characterizing uncertainty in predictions of genomic sequence-to-activity modelsPreprint
Bajwa, A., Rastogi, R., Kathail, P., Shuai, R. W., & Ioannidis, N. M. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.12.21.572730 · Cited by 7 · see citing works
“In other domains – including biological applications such as protein design – prediction uncertainty estimates have been employed to determine when to trust a model’s predictions [30–32].”
Antibody affinity engineering using antibody repertoire data and machine learningPreprint
Erlach, L., Friedensohn, S., Neumeier, D., Mason, D. M., & Reddy, S. T. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.10.632313 · Cited by 7 · see citing works
Sedgwick, R., Goertz, J., Stevens, M. M., Misener, R., & van der Wilk, M. (2024) · Biotechnology and Bioengineering · doi:10.1002/bit.28854 · Cited by 5 · see citing works
- Authors
- Hie, B., Bryson, B. D., & Berger, B.
- Affiliations
- Massachusetts Institute of Technology (US)
- Ragon Institute of MGH, MIT and Harvard (US)
- MIT Computer Science and Artificial Intelligence Laboratory (US)
- Corresponding author
- Bryson, Bryan D.
- Berger, Bonnie
- Venue
- Cell Systems
- Publisher
- Elsevier BV
- Published
- 2020-11
- Volume
- 11
- Issue
- 5
- Pages
- 461-477.e9
- ISSN
- 2405-4712 (ISSN-L)
- OpenAlex
- W3092737886
- Cited by
- 154
- References
- 125
- Citations by year
- 2026: 20 · 2025: 30 · 2024: 23 · 2023: 29 · 2022: 22 · 2021: 27 · 2020: 2
- FWCI
- 9.92
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Cell Systems (publishedVersion, cc-by-nc-nd)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
- PubMed (publishedVersion)
- DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc-nd)
- Funders
- U.S. Department of Defense
- National Institutes of Health
- National Science Foundation
- Topics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Computational Drug Discovery Methods
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
- Keywords
- uncertainty quantification, Gaussian processes, biological hypothesis generation, kinase inhibitors, Mycobacterium tuberculosis, compound screening, protein engineering, single-cell transcriptomics, drug discovery
- MeSH terms
- Machine Learning, Algorithms, Forecasting, Normal Distribution, Computational Biology, Uncertainty
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.
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
