Bioprivileged molecules — biology-derived chemical intermediates convertible by catalysis into both drop-in petrochemical replacements and novel performance-advantaged products — were identified in 2017 (Shanks & Keeling) as lacking any systematic discovery strategy: prior candidates such as muconic acid and triacetic acid lactone had been found serendipitously, and predictive links between candidate structure and end-use performance were missing. Later work substantially closed parts of this gap: computational frameworks (Zhou et al. 2019; Lopez et al. 2021) now systematically enumerate and filter candidate spaces on structural, reactivity, and thermodynamic grounds, and machine-learning quantitative structure–property models (PolyID, 2023) now connect biologically accessible monomers to predicted polymer performance with prospective experimental validation. The surviving open core is a prioritization strategy that jointly scores candidates by biological producibility, chemical diversification potential, and predicted end-use performance across application classes beyond polymers (such as antioxidants, nutraceuticals, and antimicrobials), demonstrated prospectively rather than retrospectively.

Background

A "bioprivileged molecule" is a small-molecule intermediate produced biologically from biomass (typically by engineered microbial metabolism) whose usefulness lies less in being a final product than in being a privileged starting point: chemical catalysis can convert it into many different products, including both "drop-in" replacements for existing petrochemicals (molecules identical to those already in commerce) and genuinely novel molecules whose structures are inaccessible from petroleum feedstocks and can impart improved end-use performance. The canonical examples are cis,cis-muconic acid, a dicarboxylic acid that microorganisms produce from sugars or lignin aromatics and that can be hydrogenated to adipic acid or diversified into specialty diacids and lactones, and triacetic acid lactone (TAL), a polyketide alpha-pyrone usable in polymers, antioxidants, and other applications. Shanks and Keeling (2017), writing from the NSF Engineering Research Center for Biorenewable Chemicals (CBiRC), argued that the value of this paradigm depends on which intermediates are chosen, and that the choice had so far been serendipitous. They offered only a notional heuristic for recognizing promising candidates — favoring molecules whose carbon atoms sit in distinctly different bonding environments and that carry fewer, more differentiated functional groups than the carbohydrate feedstocks — and they identified two deeper deficiencies: no rigorous methodology for systematically identifying and prioritizing candidates from the very large space of potentially bioengineerable intermediates, and little predictive structure–function connectivity between candidate molecules and their performance in end uses — application classes such as polymers, antioxidants, nutraceuticals, and antimicrobials — so that performance criteria cannot yet define desirable targets. Subsequent work changed this landscape in two ways. First, the Shanks–Broadbelt group built computational identification frameworks: Zhou et al. (2019) enumerated all 29,252 formulas in the C6HxOy space, then filtered them by structural features, reactive moieties, automated reaction-network generation, and on-the-fly thermodynamic estimation to short-list candidate bioprivileged molecules; Lopez et al. (2021) extended the approach to broader molecular space. These frameworks are chemistry-centered — they score novelty and diversification potential of the reaction chemistry, not predicted usefulness of end products. Second, for the polymer end of the problem, quantitative structure–property relationship (QSPR) models — statistical or machine-learned mappings from molecular structure to macroscopic properties — matured: Cywar et al. (2022) reviewed the emerging class of performance-advantaged bio-based polymers, and PolyID (Wilson et al. 2023) used a multi-output graph neural network (a model that learns properties directly from the molecular graph) to screen roughly 1.4 million polymers computationally reachable from biologically accessible monomers listed in the MetaCyc, MINEs, KEGG, and BiGG databases, identifying poly(ethylene terephthalate) analogues with predicted thermal and transport advantages and validating one synthesis experimentally. What no work has yet delivered is the coupling of these two strands into a single prioritization methodology that accounts for end-use performance in application classes beyond polymers — for antioxidants, nutraceuticals, and antimicrobials no predictive structure–function framework has been demonstrated and validated for bioprivileged-type candidates (the general QSAR literature on, e.g., antimicrobial or antioxidant activity does not cover these oxygen-rich, polyfunctional biomass-derived structures, nor is it coupled to candidate selection) — nor a prospective, cross-application demonstration in which the methodology, rather than luck, produces a previously unrecognized molecule that is made and shown to be performance-advantaged. The research program that posed the original problem documented this non-polymer gap directly: it targeted organic corrosion inhibitors as a non-polymer end use, and its 2023 DOE final report found no clear relationships between measured inhibition efficiency and computed molecular attributes (HOMO/LUMO energies and gaps, molecular electrostatic potential surfaces, Morgan fingerprints, surface adsorption modeling).

Problem Statement

Develop and validate a systematic, quantitative strategy for identifying and prioritizing bioprivileged molecules that replaces serendipitous candidate discovery. The strategy must jointly score candidates from the space of bioengineerable intermediates by (i) biological producibility from biomass feedstocks, (ii) chemical diversification potential — the breadth and usefulness of the product space reachable by catalytic conversion — and (iii) predicted end-use performance, extending predictive structure–function linkage beyond the polymer applications already addressed by QSPR models such as PolyID to at least one non-polymer application class, such as antioxidants, nutraceuticals, or antimicrobials, for which no predictive framework has been demonstrated for bioprivileged-derived candidates. The methodology is to identify at least one previously unrecognized candidate bioprivileged molecule — one not already identified as a bioprivileged candidate in the prior literature or present in commerce — and to specify, for its derived product, a production-and-comparison protocol by which that product would be shown to be performance-advantaged against a relevant incumbent, demonstrating that the selection — not chance — produced the value. This prospective validation is to be designed rather than executed: the resolution is a designed validation, not executed experimental results.

References

  1. Brent H. Shanks, Peter L. Keeling, "Bioprivileged molecules: creating value from biomass", Green Chemistry 19(14), 3177–3185, 2017. DOI 10.1039/c7gc00296c, https://doi.org/10.1039/c7gc00296c
  2. Xiaowei Zhou, Z. Brentzel, G. A. Kraus, P. L. Keeling, J. A. Dumesic, B. H. Shanks, L. J. Broadbelt, "Computational Framework for the Identification of Bioprivileged Molecules", ACS Sustainable Chemistry & Engineering 7(2), 2414–2428, 2019. DOI 10.1021/acssuschemeng.8b05275, https://doi.org/10.1021/acssuschemeng.8b05275
  3. Brent H. Shanks, Linda J. Broadbelt, "A Robust Strategy for Sustainable Organic Chemicals Utilizing Bioprivileged Molecules", ChemSusChem 12, 2970–2975, 2019. DOI 10.1002/cssc.201900323, https://doi.org/10.1002/cssc.201900323
  4. Wenzhao Wu, Christos T. Maravelias, "Identifying the Characteristics of Promising Renewable Replacement Chemicals", iScience 15, 136–146, 2019. DOI 10.1016/j.isci.2019.04.012, https://doi.org/10.1016/j.isci.2019.04.012
  5. Jiajie Huo, Brent H. Shanks, "Bioprivileged Molecules: Integrating Biological and Chemical Catalysis for Biomass Conversion", Annual Review of Chemical and Biomolecular Engineering 11(1), 63–85, 2020. DOI 10.1146/annurev-chembioeng-101519-121127, https://doi.org/10.1146/annurev-chembioeng-101519-121127
  6. Lauren M. Lopez, Brent H. Shanks, Linda J. Broadbelt, "Identification of bioprivileged molecules: expansion of a computational approach to broader molecular space", Molecular Systems Design & Engineering 6(6), 445–460, 2021. DOI 10.1039/D1ME00013F, https://doi.org/10.1039/D1ME00013F
  7. Robin M. Cywar, Nicholas A. Rorrer, Caroline B. Hoyt, Gregg T. Beckham, et al., "Bio-based polymers with performance-advantaged properties", Nature Reviews Materials 7, 83–103, 2022. DOI 10.1038/s41578-021-00363-3, https://doi.org/10.1038/s41578-021-00363-3
  8. Nicholas A. Rorrer, Sandra F. Notonier, Brandon C. Knott, Brenna A. Black, et al., "Production of β-ketoadipic acid from glucose in Pseudomonas putida KT2440 for use in performance-advantaged nylons", Cell Reports Physical Science 3, 100840, 2022. DOI 10.1016/j.xcrp.2022.100840, https://doi.org/10.1016/j.xcrp.2022.100840
  9. Srinivas Rangarajan, "Towards a chemistry-informed paradigm for designing molecules", Current Opinion in Chemical Engineering 35, 100717, 2022. DOI 10.1016/j.coche.2021.100717, https://doi.org/10.1016/j.coche.2021.100717
  10. A. Nolan Wilson, Peter C. St. John, et al., "PolyID: Artificial Intelligence for Discovering Performance-Advantaged and Sustainable Polymers", Macromolecules 56(21), 8547–8557, 2023. DOI 10.1021/acs.macromol.3c00994, https://doi.org/10.1021/acs.macromol.3c00994
  11. Brent H. Shanks, David Beck, Linda J. Broadbelt, Eric Cochran, George Kraus, Jim Pfaendtner, Jean-Philippe Tessonnier, "Identifying Performance Advantaged Biobased Chemicals Utilizing Bioprivileged Molecules (Final Report)", DOE/OSTI report, 2023. DOI 10.2172/2004966, https://doi.org/10.2172/2004966
  12. Liangyu Qian, Santanu Jana, Tasleem Javaid, Yunxuan Wang, Peter K. Karoki, Yunqiao Pu, Erin G. Webb, Arthur J. Ragauskas, Xiaohan Yang, Breeanna R. Urbanowicz, Joshua K. Michener, "Biochemical approaches for synthesis of performance-advantaged polymers from lignocellulosic biomass", Green Chemistry, 2026. DOI 10.1039/d6gc01235c, https://doi.org/10.1039/d6gc01235c