[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122841-en":3,"doc-seo-122841-105":30,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},122841,4810365810221,"Aurora","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Electronic structure at coarse-grained resolutions from supervised machine learning - Supplementary Materials","Supplementary materials provide supporting figures and tables for a supervised machine learning framework that predicts electronic structure using coarse-grained resolutions. The document reports ANN-ECG model performance and temperature transferability via 2D histogram analyses across rigid and flexible datasets, including HOMO energy distributions. It also details tight-binding fitting parameters, an ANN hyperparameter grid search for layers, neurons, epochs, and regularization, and presents coarse-grained atomic mapping for 3MT monomers used in main-text Figure 4.","[advances.sciencemag.org/cgi/content/full/5/3/eaav1190/DC1](advances.sciencemag.org/cgi/content/full/5/3/eaav1190/DC1)  \nSupplementary Materials for  \nElectronic structure at coarse-grained resolutions from supervised machine learning  \nNicholas E. Jackson, Alec S. Bowen, Lucas W. Antony, Michael A. Webb,  \nVenkatram Vishwanath, Juan J. de Pablo*  \n*Corresponding author. Email: [depablo@uchicago.edu](depablo@uchicago.edu)  \n[Published 22 March 2019](Published 22 March 2019), Sci. Adv. 5, eaav1190 (2019)  \nDOI: 10. 1126/sciadv.aav1190  \nThis PDF file includes:  \nFig. S1 . Temperature transferability of the ANN-ECG model.  \nFig. S2. ANN-ECG performance versus training data aize for 500 K/rigid dataset.  \nFig. S3 . Distribution of HOMO energy levels for 300 K/flexible and 300 K/rigid datasets.  \nFig. S4 . Atomic numbering scheme used for each 3MT monomer.  \nFig. S5. Delta–machine learning fitting results for ANN-ECG using 300 K/rigid dataset.  \nFig. S6 . Application of ANN-ECG to conjugated copolymer PTB7 and non-fullerene acceptor TPB.  \nFig. S7 . ANN-ECG results for the HOMO-5→HOMO energy levels ofS3MT using 300 K/rigid dataset computed at the BP86/6-31G* level of theory.  \nTable S1 . Hyperparameter optimization for ANN layers and neurons.  \nTable S2. Hyperparameter optimization for number of training epochs.  \nTable S3 . Results using ANN-ECG and a systematic coarse-graining strategy.  \nANN-ECG temperature transferability  \nFig. S1. Temperature transferability of the ANN-ECG model. 2D histograms ofANN-ECG performance (A) trained on 300 K/rigid applied to 500 K/rigid and (B) trained on 500 K/rigid applied to 300 K/rigid. Colorbar denotes the probability distribution of predicted HOMO energy levels, and the inset shows the prediction in the interval of the highest-energy HOMO.  \nANN-[ECG performance vs. training](ECG performance vs. training) data size  \nFig. S2. ANN-ECG performance versus training data aize for 500 K/rigid dataset. Plot of ANN-ECG performance vs. size of training set for the 500 K/rigid data set of S3MT. RMSE (green) and r2 (blue) error bars obtained via 5-fold cross-validation applied to a held-out 1,000 configuration validation data set. Error bars represent one standard deviation.  \nTight-binding model fitting parameters  \nTemp/Condition 300K/rigid 500K/rigid 300K/flex 500K/flex  \n􀁈 (eV)  \n-8.0824  \n-8.0605  \n-8.0418  \n-8.0026  \nti,i+1 (eV)  \n0.902 0.905 0.890  \n0.888  \nModified 2-Band tight-binding Hamiltonian results  \nTo explore the accuracy of more complicated tight-binding models, we applied a simplex fitting procedure for a two-band tight-binding model with distinct “middle” and “end” sites. The two bands correspond to the HOMO and HOMO-1 energies for each thiophene monomer. This model includes a total of 7 fitting parameters (4 energies – HOMOmiddle, HOMOend, HOMO-  \n1middle, HOMO-1end, 3 couplings – HOMOi-HOMOi+1, HOMOi-HOMO-1i+1,HOMO-1i-HOMO- 1i+1). All couplings were assumed to be proportional to the cosine of the dihedral angle between neighboring monomers. This tight-binding Hamiltonian was regressed to the 300K/rigid data set. The obtained performance was quantitatively similar to that derived from the simple 1-band tight-binding model. These results obtained a RMSE 54.1 +/ -0.9 meV of and a r2 of 0.784 +/ - 0.001  \nExample HOMO energy distributions from ZINDO/S  \nFig. S3. Distribution of HOMO energy levels for 300 K/flexible and 300 K/rigid datasets.  \nHyperparameter optimizations  \nWe performed a hyperparameter grid search of the number of layers in the ANN, as well as the number of neurons within each layer. Improvements were not observed for more than 2 hidden layers. Performance estimates occurred for 10,000 epochs with 1,000 batch size using 5-fold cross-validated RMSE and r2 on the 300K/rigid data set using the 3-bead/3MT orthogonal coordinate system monomer mapping.  \nTable S1. Hyperparameter optimization for ANN layers and neurons. We also optimized the number of training epochs as a hy","cbCaisprq23AljwS","https://ap.wps.com/l/cbCaisprq23AljwS","pdf",983792,1,19,"English","en",105,"# Supplementary materials overview\n## Figures\n## Tables\n## ANN-ECG temperature transferability\n## ANN-ECG performance vs. training data size\n## Tight-binding model fitting parameters\n## Modified two-band tight-binding results\n## Example HOMO energy distributions from ZINDO/S\n## Hyperparameter optimizations\n## Coarse-grained mappings for main-text Figure 4","[{\"question\":\"What does the ANN-ECG temperature transferability analysis compare?\",\"answer\":\"It compares ANN-ECG performance when trained on one temperature dataset (e.g., 300 K/rigid or 500 K/rigid) and applied to the other, using 2D histograms of predicted HOMO energy distributions.\"},{\"question\":\"Which ANN hyperparameters were optimized?\",\"answer\":\"The materials describe grid search over the number of ANN layers and neurons, and separate testing of the number of training epochs as a hyperparameter, with cross-validated RMSE and r2 used for performance estimates.\"},{\"question\":\"How are coarse-grained atomic groupings defined for each 3MT monomer?\",\"answer\":\"Coarse-grained mappings are generated by graph-based coarse-graining with spectral grouping iterations, and the document lists grouping schemes for multiple coarse-grained resolution levels (from atomistic to more merged united-atom representations).\"}]","Electronic structure at coarse-grained resolutions from supervised machine learning - Supplementary Materials | PDF",1785813198,48,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"electronic-structure-at-coarse-grained-resolutions-from-supervised-machine-learning-supplementary-materials","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/electronic-structure-at-coarse-grained-resolutions-from-supervised-machine-learning-supplementary-materials/122841/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05","2026-08-04",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"What does the ANN-ECG temperature transferability analysis compare?","Question",{"text":76,"@type":77},"It compares ANN-ECG performance when trained on one temperature dataset (e.g., 300 K/rigid or 500 K/rigid) and applied to the other, using 2D histograms of predicted HOMO energy distributions.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"Which ANN hyperparameters were optimized?",{"text":81,"@type":77},"The materials describe grid search over the number of ANN layers and neurons, and separate testing of the number of training epochs as a hyperparameter, with cross-validated RMSE and r2 used for performance estimates.",{"name":83,"@type":74,"acceptedAnswer":84},"How are coarse-grained atomic groupings defined for each 3MT monomer?",{"text":85,"@type":77},"Coarse-grained mappings are generated by graph-based coarse-graining with spectral grouping iterations, and the document lists grouping schemes for multiple coarse-grained resolution levels (from atomistic to more merged united-atom representations).","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":107,"slug":138},"General","general"]