{"date_created":"2025-01-12T23:04:00Z","oa":1,"date_published":"2025-01-02T00:00:00Z","file_date_updated":"2025-01-14T06:59:25Z","citation":{"chicago":"Wild, Romina, Felix Wodaczek, Vittorio Del Tatto, Bingqing Cheng, and Alessandro Laio. “Automatic Feature Selection and Weighting in Molecular Systems Using Differentiable Information Imbalance.” <i>Nature Communications</i>. Springer Nature, 2025. <a href=\"https://doi.org/10.1038/s41467-024-55449-7\">https://doi.org/10.1038/s41467-024-55449-7</a>.","ista":"Wild R, Wodaczek F, Del Tatto V, Cheng B, Laio A. 2025. Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance. Nature Communications. 16, 270.","ieee":"R. Wild, F. Wodaczek, V. Del Tatto, B. Cheng, and A. Laio, “Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance,” <i>Nature Communications</i>, vol. 16. Springer Nature, 2025.","short":"R. Wild, F. Wodaczek, V. Del Tatto, B. Cheng, A. Laio, Nature Communications 16 (2025).","mla":"Wild, Romina, et al. “Automatic Feature Selection and Weighting in Molecular Systems Using Differentiable Information Imbalance.” <i>Nature Communications</i>, vol. 16, 270, Springer Nature, 2025, doi:<a href=\"https://doi.org/10.1038/s41467-024-55449-7\">10.1038/s41467-024-55449-7</a>.","ama":"Wild R, Wodaczek F, Del Tatto V, Cheng B, Laio A. Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance. <i>Nature Communications</i>. 2025;16. doi:<a href=\"https://doi.org/10.1038/s41467-024-55449-7\">10.1038/s41467-024-55449-7</a>","apa":"Wild, R., Wodaczek, F., Del Tatto, V., Cheng, B., &#38; Laio, A. (2025). Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance. <i>Nature Communications</i>. Springer Nature. <a href=\"https://doi.org/10.1038/s41467-024-55449-7\">https://doi.org/10.1038/s41467-024-55449-7</a>"},"volume":16,"publication_identifier":{"eissn":["2041-1723"]},"file":[{"access_level":"open_access","relation":"main_file","content_type":"application/pdf","success":1,"checksum":"b3d0f3568d9a87c494cf231a5324029a","file_id":"18846","date_updated":"2025-01-14T06:59:25Z","creator":"dernst","file_name":"2025_NatureComm_Wild.pdf","file_size":1216738,"date_created":"2025-01-14T06:59:25Z"}],"acknowledgement":"The authors thank Dr. Matteo Carli for providing the CLN025 replica exchange MD trajectory and Matteo Allione for the fruitful discussions connected with the idea of the linear scaling estimator. This work was partially funded by NextGenerationEU through the Italian National Centre for HPC, Big Data, and Quantum Computing (Grant No. CN00000013 received by A.L.). A.L. also acknowledges financial support by the region Friuli Venezia Giulia (project F53C22001770002 received by A.L.).","OA_place":"publisher","author":[{"full_name":"Wild, Romina","first_name":"Romina","last_name":"Wild"},{"full_name":"Wodaczek, Felix","first_name":"Felix","id":"8b4b6a9f-32b0-11ee-9fa8-bbe85e26258e","last_name":"Wodaczek","orcid":"0009-0000-1457-795X"},{"last_name":"Del Tatto","full_name":"Del Tatto, Vittorio","first_name":"Vittorio"},{"first_name":"Bingqing","full_name":"Cheng, Bingqing","last_name":"Cheng","id":"cbe3cda4-d82c-11eb-8dc7-8ff94289fcc9","orcid":"0000-0002-3584-9632"},{"last_name":"Laio","full_name":"Laio, Alessandro","first_name":"Alessandro"}],"article_type":"original","pmid":1,"publication":"Nature Communications","intvolume":"        16","OA_type":"gold","DOAJ_listed":"1","language":[{"iso":"eng"}],"year":"2025","publisher":"Springer Nature","article_processing_charge":"Yes","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","doi":"10.1038/s41467-024-55449-7","date_updated":"2025-01-14T07:00:17Z","external_id":{"pmid":["39747013"]},"ddc":["570"],"abstract":[{"lang":"eng","text":"Feature selection is essential in the analysis of molecular systems and many other fields, but several uncertainties remain: What is the optimal number of features for a simplified, interpretable model that retains essential information? How should features with different units be aligned, and how should their relative importance be weighted? Here, we introduce the Differentiable Information Imbalance (DII), an automated method to rank information content between sets of features. Using distances in a ground truth feature space, DII identifies a low-dimensional subset of features that best preserves these relationships. Each feature is scaled by a weight, which is optimized by minimizing the DII through gradient descent. This allows simultaneously performing unit alignment and relative importance scaling, while preserving interpretability. DII can also produce sparse solutions and determine the optimal size of the reduced feature space. We demonstrate the usefulness of this approach on two benchmark molecular problems: (1) identifying collective variables that describe conformations of a biomolecule, and (2) selecting features for training a machine-learning force field. These results show the potential of DII in addressing feature selection challenges and optimizing dimensionality in various applications. The method is available in the Python library DADApy."}],"quality_controlled":"1","has_accepted_license":"1","scopus_import":"1","department":[{"_id":"AnSa"},{"_id":"BiCh"}],"title":"Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance","oa_version":"Published Version","status":"public","month":"01","type":"journal_article","day":"02","publication_status":"published","article_number":"270","tmp":{"image":"/images/cc_by_nc_nd.png","name":"Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)","legal_code_url":"https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode","short":"CC BY-NC-ND (4.0)"},"_id":"18820"}