182 Expanding Geology’s Horizons: Geoinformatics, Open Science, and Open Data (Posters)

This session highlights the development of new methodologies for collecting, sharing, and analyzing geological datasets and the advancement of FAIR (findable, accessible, interoperable, and reusable) data practices.

Tuesday, 21 October 2025
8:00 AM - 5:30 PM
HBGCC, Hall 1

Characterizing Earth’s Earliest Life using machine learning

Paper number: 182-2
Program: Technical Sessions
Time: 8:00 AM – 8:00 AM

Authors

Abstract

The study of early life offers complex challenges and rich opportunities that can be effectively addressed through modern machine learning techniques. In particular, classifying ancient samples containing highly altered organic residues requires the development of customized data-driven approaches that can extract biosignatures from complex geochemical patterns. 

We have developed supervised machine learning pipelines that include data processing, feature selection, model interpretation, optimization and rigorous validation to identify biosignatures from diverse molecular data. We start with a curated set of 406 samples, from modern biota to thermally altered organic matter. We have evaluated multiple ensemble learning methods for distinguishing various groups of samples including biotic vs abiotic, photosynthetic vs non-photosynthetic and plants vs animals. Random forests emerged as the most accurate, robust and interpretable classifier achieving over 93% accuracy across our desired classification tasks. 

In this contribution, we present the development of our machine learning model to identify molecular biosignatures in some of the oldest samples on Earth. We also discuss the key considerations on working with ancient samples that produce complex spectra and discuss some key findings made by our team. For example, we found that some organic suites altered by temperatures greater than 400C, such as those in coal slag and charcoal, have had most of their biotic/photosynthetic signatures obliterated. Nevertheless, we present evidence for biotic molecular assemblages in Paleoarchean sedimentary rocks (3.33 Ga) and for photosynthetic life in Neoarchean sedimentary rocks (2.52 Ga)—results consistent with morphological and isotopic data. Lastly, we also present our future plans for a more systematic exploration of ancient samples through data-driven and informatics methods.

Geological Society of America Abstracts with Programs. Vol. 57, No. 6

doi: 10.1130/abs/2025AM-414537

© Copyright 2025 The Geological Society of America (GSA), all rights reserved. Permission is hereby granted to the author(s) of this abstract to reproduce and distribute it freely, for noncommercial purposes. Permission is hereby granted to any individual scientist to download a single copy of this electronic file and reproduce up to 20 paper copies for noncommercial purposes advancing science and education, including classroom use, providing all reproductions include the complete content shown here, including the author information. All other forms of reproduction and/or transmittal are prohibited without written permission from GSA Copyright Permissions.