OpenLB
Open Library of Bioscience
Advanced Search
Proceedings of the National Academy of Sciences of the United States of America
10 10, 2023;120 (41): e2307149120.

A robust, agnostic molecular biosignature based on machine learning.

H James Cleaves, Grethe Hystad, Anirudh Prabhu, Michael L Wong, George D Cody, Sophia Economon, Robert M Hazen
Author Information
  1. H James Cleaves: Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC 20015. ORCID
  2. Grethe Hystad: Department of Mathematics and Statistics, Purdue University Northwest, Hammond, IN 46323. ORCID
  3. Anirudh Prabhu: Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC 20015. ORCID
  4. Michael L Wong: Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC 20015. ORCID
  5. George D Cody: Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC 20015. ORCID
  6. Sophia Economon: Department of Earth and Planetary Sciences, Johns Hopkins University, Baltimore, MD 21218.
  7. Robert M Hazen: Earth and Planets Laboratory, Carnegie Institution for Science, Washington, DC 20015. ORCID
PMID: 37748080 DOI: 10.1073/pnas.2307149120

Abstract

The search for definitive biosignatures-unambiguous markers of past or present life-is a central goal of paleobiology and astrobiology. We used pyrolysis-gas chromatography coupled to mass spectrometry to analyze chemically disparate samples, including living cells, geologically processed fossil organic material, carbon-rich meteorites, and laboratory-synthesized organic compounds and mixtures. Data from each sample were employed as training and test subsets for machine-learning methods, which resulted in a model that can identify the biogenicity of both contemporary and ancient geologically processed samples with ~90% accuracy. These machine-learning methods do not rely on precise compound identification: Rather, the relational aspects of chromatographic and mass peaks provide the needed information, which underscores this method's utility for detecting alien biology.

Keywords

biosignatures carbonaceous meteorites machine learning organic chemistry taphonomy

References

  1. Nature. 2001 Mar 1;410(6824):77-81 [PMID: 11242044]
  2. PLoS Biol. 2004 Sep;2(9):E302 [PMID: 15367939]
  3. Adv Carbohydr Chem. 1959;14:63-134 [PMID: 13820038]
  4. Chem Sci. 2022 Mar 23;13(17):4838-4853 [PMID: 35655880]
  5. Philos Trans A Math Phys Eng Sci. 2017 Dec 28;375(2109): [PMID: 29133442]
  6. Astrobiology. 2022 Feb;22(2):143-157 [PMID: 35021862]
  7. Astrobiology. 2018 Jul;18(7):915-922 [PMID: 29634318]
  8. Nat Cell Biol. 1999 May;1(1):E13-5 [PMID: 10559875]
  9. Astrobiology. 2012 Jan;12(1):29-39 [PMID: 22204400]
  10. Anal Chem. 2012 Apr 3;84(7):3287-95 [PMID: 22414024]
  11. Nat Commun. 2021 May 24;12(1):3033 [PMID: 34031398]
  12. Nature. 2004 Dec 16;432(7019):824-8 [PMID: 15602547]
  13. Proc Natl Acad Sci U S A. 2023 Oct 10;120(41):e2307149120 [PMID: 37748080]
  14. Life (Basel). 2021 Mar 12;11(3): [PMID: 33809046]
  15. J R Soc Interface. 2023 Feb;20(199):20220810 [PMID: 36751931]
  16. Astrobiology. 2023 Jan;23(1):60-75 [PMID: 36454287]
  17. Appl Microbiol. 1970 Oct;20(4):567-72 [PMID: 16349890]
  18. Nature. 1965 Aug 7;207(997):568-70 [PMID: 5883628]
  19. Nature. 1961 Jul 8;191:144-8 [PMID: 13771349]
  20. Astrobiology. 2017 Oct;17(10):1022-1052 [PMID: 28443722]
  21. Astrobiology. 2014 Jun;14(6):541-9 [PMID: 24735484]
  22. Science. 2015 Jan 23;347(6220):415-7 [PMID: 25515120]
  23. Entropy (Basel). 2022 Sep 28;24(10): [PMID: 37420403]
  24. Astrobiology. 2018 Jun;18(6):779-824 [PMID: 29938538]
  25. Bull Math Biophys. 1969 Mar;31(1):159-98 [PMID: 5779774]
  26. J Geophys Res. 1999 Nov 25;104(E11):26977-95 [PMID: 11543200]
  27. J Mol Evol. 2011 Mar;72(3):283-95 [PMID: 21253717]
  28. Proc Natl Acad Sci U S A. 2007 Jun 19;104(25):10310-3 [PMID: 17548829]
  29. Bioinformatics. 2012 Sep 1;28(17):2270-1 [PMID: 22796955]
  30. Annu Rev Earth Planet Sci. 1999;27:313-58 [PMID: 11543060]
  31. Astrobiology. 2017 Sep;17(9):902-912 [PMID: 28915087]
  32. Astrobiology. 2017 Mar;17(3):216-230 [PMID: 28323483]
  33. Astrobiology. 2022 Sep;22(9):1099-1115 [PMID: 35749703]
  34. Nature. 1970 Dec 5;228(5275):923-6 [PMID: 5482102]
  35. Science. 2000 Aug 4;289(5480):756-9 [PMID: 10926533]
  36. Astrobiology. 2019 Sep;19(9):1075-1102 [PMID: 31335163]
  37. J Am Soc Mass Spectrom. 2023 Aug 2;34(8):1584-1592 [PMID: 37390315]
  38. Astrobiology. 2021 Jul;21(7):866-892 [PMID: 34042490]
  39. Proc Natl Acad Sci U S A. 2010 Feb 16;107(7):2763-8 [PMID: 20160129]
  40. Astrobiology. 2018 Jun;18(6):663-708 [PMID: 29727196]

MeSH Term

Humans
Carbon
Emigrants and Immigrants
Exobiology
Fossils
Machine Learning

Chemicals

Carbon
Journal Article Research Support, U.S. Gov't, Non-P.H.S. Research Support, Non-U.S. Gov't
Article has an altmetric score of 1968

Word Cloud

Created with Highcharts 10.0.0organicmasssamplesgeologicallyprocessedmeteoritesmachine-learningmethodsmachinelearningsearchdefinitivebiosignatures-unambiguousmarkerspastpresentlife-iscentralgoalpaleobiologyastrobiologyusedpyrolysis-gaschromatographycoupledspectrometryanalyzechemicallydisparateincludinglivingcellsfossilmaterialcarbon-richlaboratory-synthesizedcompoundsmixturesDatasampleemployedtrainingtestsubsetsresultedmodelcanidentifybiogenicitycontemporaryancient~90%accuracyrelyprecisecompoundidentification:Ratherrelationalaspectschromatographicpeaksprovideneededinformationunderscoresmethod'sutilitydetectingalienbiologyrobustagnosticmolecularbiosignaturebasedbiosignaturescarbonaceouschemistrytaphonomy

Similar Articles

Cited By