journal article·2026-05Open access

BACH, a Bayesian Optimization Protocol for Accurate Coarse‐Grained Parameterization of Organic Liquids

Janak Prabhu, Akhil Pratap Singh, Pablo Campomanes, Stefano Vanni

Advanced Functional Materials· Volume 36· Issue 38

Abstract

ABSTRACT Coarse‐grained (CG) molecular dynamics simulations are essential tools to investigate large‐scale soft‐matter and biomolecular systems, providing information often inaccessible through experimental methods. However, the accuracy of CG force fields critically depends on the underlying parameterization strategy, which is traditionally labor‐intensive and must balance computational efficiency and physical fidelity. Here, we present BACH, a fully automated Bayesian optimization (BO) protocol for the parameterization of nonbonded interactions in CG force fields. Using experimental thermophysical data—including density, surface and interfacial tension, enthalpy of vaporization, and isothermal compressibility—we apply the protocol to a broad range of liquids, spanning linear, branched, and unsaturated hydrocarbons, esters, triglycerides, and water. BACH improves upon MARTINI by approximately 10%–15% in density, 5%–20% in surface tension, 20% in enthalpy of vaporization, and 40% in isothermal compressibility. Compared with SPICA, similar accuracy is obtained for density and surface tension while enthalpy of vaporization and compressibility improve by 5%–20% and ∼40%, respectively. BACH transferability is demonstrated across temperatures, chemical families, and properties in independent validation sets. These results indicate that the proposed BO framework, if supplied with appropriate experimental data, can be systematically extended to more complex chemical spaces, including biomolecular and functional or bio‐inspired materials.
Source: crossref · Quality: baseline

Research context

OpenAlex topics: Protein Structure and Dynamics · Machine Learning in Materials Science · Block Copolymer Self-Assembly

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

Department of Biology University of Fribourg Fribourg Switzerland

National Center of Competence in Research Bio‐Inspired Materials University of Fribourg Fribourg Switzerland

Funding

European Research Council · Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung

References (97)

Show source-provided references
1. Frenkel D. · 2023
2. 10.1093/oso/9780198803195.001.000110.1093/oso/9780198803195.001.0001 ↗
3. 10.1119/10.000065410.1119/10.0000654 ↗
4. 10.1016/j.neuron.2018.08.01110.1016/j.neuron.2018.08.011 ↗
5. 10.1002/adfm.20241370310.1002/adfm.202413703 ↗
6. 10.1038/s41586‐023‐06050‐310.1038/s41586‐023‐06050‐3 ↗
7. 10.1038/nsb0902‐64610.1038/nsb0902‐646 ↗
8. 10.1039/D2NR02824G10.1039/D2NR02824G ↗
9. 10.1021/acs.chemrev.5b0074410.1021/acs.chemrev.5b00744 ↗
10. 10.1039/D1CP03276C10.1039/D1CP03276C ↗
11. 10.1002/adfm.20200696910.1002/adfm.202006969 ↗
12. 10.1021/acs.jcim.4c0152010.1021/acs.jcim.4c01520 ↗
13. 10.1073/pnas.20242739910.1073/pnas.202427399 ↗
14. 10.1016/0021‐9991(77)90121‐810.1016/0021‐9991(77)90121‐8 ↗
15. 10.1038/s41592‐021‐01098‐310.1038/s41592‐021‐01098‐3 ↗
16. 10.1080/0892702060105405010.1080/08927020601054050 ↗
17. 10.1021/acs.jctc.3c0073310.1021/acs.jctc.3c00733 ↗
18. 10.1021/acs.jctc.8b0098710.1021/acs.jctc.8b00987 ↗
19. 10.1021/acs.chemrev.6b0016310.1021/acs.chemrev.6b00163 ↗
20. 10.1021/acs.jctc.1c0120710.1021/acs.jctc.1c01207 ↗
21. 10.1002/adma.20200863510.1002/adma.202008635 ↗
22. 10.1021/acs.jpcb.2c0873110.1021/acs.jpcb.2c08731 ↗
23. 10.1002/wcms.7002210.1002/wcms.70022 ↗
24. 10.3389/frsfm.2024.136106610.3389/frsfm.2024.1361066 ↗
25. 10.1021/acs.jctc.2c0064310.1021/acs.jctc.2c00643 ↗
26. 10.1021/acs.jpcb.7b0999310.1021/acs.jpcb.7b09993 ↗
27. 10.1081/SMTS‐12001674610.1081/SMTS‐120016746 ↗
28. 10.1021/jp010238p10.1021/jp010238p ↗
29. 10.1021/jp044629q10.1021/jp044629q ↗
30. 10.1063/5.012453810.1063/5.0124538 ↗
31. 10.1021/acscentsci.8b0091310.1021/acscentsci.8b00913 ↗
32. 10.1063/5.002613310.1063/5.0026133 ↗
33. 10.1021/acs.jctc.0c0134310.1021/acs.jctc.0c01343 ↗
34. 10.1021/acs.jctc.3c0001610.1021/acs.jctc.3c00016 ↗
35. 10.1021/acs.jctc.3c0070210.1021/acs.jctc.3c00702 ↗
36. 10.1016/j.sbi.2019.12.00510.1016/j.sbi.2019.12.005 ↗
37. 10.1039/C8CP05889J10.1039/C8CP05889J ↗
38. 10.1063/5.007904410.1063/5.0079044 ↗
39. 10.1021/jp910720610.1021/jp9107206 ↗
40. 10.1021/acs.jctc.3c0101610.1021/acs.jctc.3c01016 ↗
41. 10.1021/jp071097f10.1021/jp071097f ↗
42. 10.1021/ct700324x10.1021/ct700324x ↗
43. 10.1039/c1sm05173c10.1039/c1sm05173c ↗
44. 10.1016/j.cpc.2009.10.02410.1016/j.cpc.2009.10.024 ↗
45. 10.1002/jcc.2613110.1002/jcc.26131 ↗
46. 10.1021/jp409545x10.1021/jp409545x ↗
47. 10.1016/S0010‐4655(02)00562‐310.1016/S0010‐4655(02)00562‐3 ↗
48. 10.1021/acs.jpcb.8b0382210.1021/acs.jpcb.8b03822 ↗
49. 10.1039/C9ME00173E10.1039/C9ME00173E ↗
50. 10.1038/s41467-018-08222-610.1038/s41467-018-08222-6 ↗
51. 10.1162/10636560175019039810.1162/106365601750190398 ↗
52. 10.1039/D2DD00138A10.1039/D2DD00138A ↗
53. 10.1021/acs.jcim.1c0044810.1021/acs.jcim.1c00448 ↗
54. 10.1002/mats.20250008110.1002/mats.202500081 ↗
55. 10.1080/00268976.2020.178557110.1080/00268976.2020.1785571 ↗
56. 10.1021/acs.jcim.3c0018610.1021/acs.jcim.3c00186 ↗
57. 10.1021/acs.jctc.2c0097510.1021/acs.jctc.2c00975 ↗
58. 10.1080/00268976.2010.50319710.1080/00268976.2010.503197 ↗
59. 10.1021/acs.jpcb.7b1054210.1021/acs.jpcb.7b10542 ↗
60. 10.1371/journal.pcbi.100081010.1371/journal.pcbi.1000810 ↗
61. 10.1021/acs.jctc.9b0094610.1021/acs.jctc.9b00946 ↗
62. 10.1021/ct200142z10.1021/ct200142z ↗
63. 10.1021/acs.jctc.1c0126810.1021/acs.jctc.1c01268 ↗
64. 10.1016/j.bpj.2017.02.03210.1016/j.bpj.2017.02.032 ↗
65. 10.1039/b808701f10.1039/b808701f ↗
66. 10.1021/acs.jced.9b0052510.1021/acs.jced.9b00525 ↗
67. 10.1021/je000038r10.1021/je000038r ↗
68. 10.1021/je010288n10.1021/je010288n ↗
69. 10.1021/accountsmr.2c0012910.1021/accountsmr.2c00129 ↗
70. 10.1063/1.174715210.1063/1.1747152 ↗
71. 10.1021/acs.jctc.7b0080010.1021/acs.jctc.7b00800 ↗
72. 10.7554/eLife.6288610.7554/eLife.62886 ↗
73. 10.1016/j.abb.2008.02.03610.1016/j.abb.2008.02.036 ↗
74. 10.1016/j.bpr.2021.10003410.1016/j.bpr.2021.100034 ↗
75. 10.1021/acs.jpcb.1c0355910.1021/acs.jpcb.1c03559 ↗
76. 10.1007/s11746‐017‐3030‐y10.1007/s11746‐017‐3030‐y ↗
77. 10.1007/s11483‐016‐9466‐z10.1007/s11483‐016‐9466‐z ↗
78. 10.1021/j100367a07710.1021/j100367a077 ↗
79. 10.1063/1.167990310.1063/1.1679903 ↗
80. 10.1063/5.024933310.1063/5.0249333 ↗
81. 10.1016/j.jfoodeng.2012.01.02810.1016/j.jfoodeng.2012.01.028 ↗
82. 10.1021/jp101759q10.1021/jp101759q ↗
83. 10.1021/je60044a00310.1021/je60044a003 ↗
84. Linstrom P. J. · 2014
85. Lide D. R. · 2006
86. Yaws C. L. · 1999
87. Yaws C. L. · 2014
88. Nogueira F. · 2014
89. 10.1002/andp.1903316080210.1002/andp.19033160802 ↗
90. 10.1093/bioinformatics/btt05510.1093/bioinformatics/btt055 ↗
91. 10.1016/j.softx.2015.06.00110.1016/j.softx.2015.06.001 ↗
92. 10.1007/978-3-319-15976-8_110.1007/978-3-319-15976-8_1 ↗
93. 10.1063/1.240842010.1063/1.2408420 ↗
94. 10.1063/1.32869310.1063/1.328693 ↗
95. 10.1103/PhysRev.159.9810.1103/PhysRev.159.98 ↗
96. 10.1063/1.174778210.1063/1.1747782 ↗
97. 10.1080/0361092890883012710.1080/03610928908830127 ↗