Algorithmic Classification of Raman Spectra Biosignatures: Improving Life Detection Confidence

Tao Sheng; Aarya Mishra; Jesse B Murray; Sunanda Sharma; Diana Gentry

NASA NTRS · 20210020822

Algorithmic Classification of Raman Spectra Biosignatures: Improving Life Detection Confidence

Tao Sheng·Aarya Mishra·Jesse B Murray·Sunanda Sharma·Diana Gentry

Abstract

“Agnostic” biosignatures – indicators of life (or the absence of life), independent of a particular biochemistry – are increasingly considered a high standard for life detection. The Ladder of Life Detection (2018) called for investigating how combinations of independent and different potential biosignatures affect confidence. To address this gap, statistical classification of elemental abundances, isotopic fractionation, and reflectance spectroscopy (VNIR) has been implemented. Raman spectroscopy, highly desirable due to its wide availability, has the potential to improve this predictive power. This work implemented biosignature classification algorithms on Raman data alone, in preparation for combination with the other data types. Raman spectroscopy data was collected from published databases and papers as part of a manually curated dataset of “indicative” and “non-indicative of life” samples. These currently include 61 non-indicative samples (meteorites, magnetite); 3 indicative living samples (bacteria); 20 indicative non-living samples (chalk, bone); and 12 indicative mixed (with non-indicative material) samples (soil, microbial mats). Laboratory work is ongoing to characterize additional samples, particularly a greater breadth of mixed systems. Spectra were interpolated, filtered with the Savitzsky-Golay filter, and de-noised. For a preliminary examination, agnostic features were manually extracted including mean intensity, number of peaks, and mean peak width. Different peak prominences and filtering polynomials were used to refine features. Classification algorithms were implemented: k-nearest neighbors (KNN), logistic regression (LR), linear support vector machines (SVM), random forest (RF), Gaussian naïve bayes (GNB). Lastly, Monte Carlo simulations on 1,000 50%-train-test-splits were used to validate classification performance and feature significance. The preliminary feature set achieved its highest AUC of 0.52 with LR, with no strongly discriminatory features. Work to improve feature extraction, such as through deep learning with back propagation, is planned. In future work, the Raman data will be combined with the other data types, and potentially new data types such as enantiomeric excess. This project was partially supported through the NASA Ames Project EXcellence (APEX) incubator program.

Explore related subjects

Astrobiology biosignatures life-detection machine learning algorithms Raman spectroscopy data analysis

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Tao Sheng, Aarya Mishra, Jesse B Murray, Sunanda Sharma, Diana Gentry. Algorithmic Classification of Raman Spectra Biosignatures: Improving Life Detection Confidence. https://ntrs.nasa.gov/citations/20210020822

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

NASA NTRS2025

Laminae as Structural Biosignatures in NASA’s Life Detection Knowledge Base

Laminae are millimeter-scale features in rocks created by physiochemical processes that can be influenced by the presence and activities of communities of organisms that occur as biofilms and microbial mats. The structure and composition of laminae reflect the processes involved in their formation and can be preserved in the rock record over geologic time, however diagenetic and metamorphic alteration can lead to the loss of primary information and confusion over the interpretation of their origins. As potential records of ancient life, laminae can preserve evidence of microbial activity over billions of years of Earth’s history. On planetary bodies beyond Earth, laminae in sedimentary rocks are the most common sedimentary structure and represent significant features of interest that can record habitable conditions (e.g., the presence of liquid water) at the time of their formation. Here we review the significance of laminae as targets for astrobiological exploration. We discuss common mechanisms by which laminae form in natural environments on Earth, present arguments and evidence for laminae as biosignatures, and describe how such information is incorporated into the NASA Life Detection Knowledge Base.

Astrobiology↗

NASA NTRS2024

Multiple Habitable Phases on Outer Exosolar Worlds

As stars evolve to higher luminosities during first ascension of the giant branch, previously frozen terrestrial worlds may thaw and host liquid water on their surfaces. Eventually these outer worlds again become uninhabitable due to receiving too much incident light and their water inventory evaporating. Solar-mass stars experience a sudden decrease in luminosity entering the horizontal branch, which could result in a secondary habitable phase for their outer worlds. The outer worlds' time with habitable surface climates is key in evaluating the possibility of extraterrestrial life arising. The times inside the habitable zone (TIHZ) are calculated for outer worlds orbiting between 5 and 45 au around a Sun-like star. By comparing the TIHZ to time estimates for life to arise on Earth, we evaluate whether such outer worlds are promising candidates in the search for extraterrestrial life. We use two different solar evolution models (PARSEC and Dartmouth) and both optimistic and conservative habitable zone (HZ) definitions. Multiple habitable phases are found for each outer world. Outer worlds with orbits as large as Saturn are found to have a secondary habitable phase which exceeds the first in duration. Generally, the time inside the HZ is found to decrease almost monotonically with orbiting distance. Water loss is calculated after the first habitable phase to determine whether a secondary habitable phase is possible. For all orbiting distances the water loss is insufficient to deplete a water inventory equivalent to that of many moons in the outer solar system.

Astrobiology↗

NASA NTRS2022

Co-leveraging Scientific Advances in Space Biology and Astrobiology Towards Achieving NASA’s Life Science Objectives

Executive Summary: Distinct lines of scientific inquiry drives the separation of NASA’s fundamental life science research into Space Biology and Astrobiology. This division developed as a way to place life scientists alongside experts in the physical constraints that define the acclimation, adaptation and evolution of biology systems relevant to their respective subjects. For astrobiology, integration with disciplines such as geology, geochemistry, astronomy, planetary science, etc., enables a comprehensive assessment of the physical environment and its co-evolution with biological processes. Space Biology’s co-location with Physical Sciences places life science researchers adjacent to experts in the physical phenomena associated with microgravity and spaceflight, enabling an understanding of how the spaceflight environment affects biological systems. Despite this separation, aspects of both disciplines have converged on a similar, fundamental objective: to describe and understand the dynamics of complex living communities in the contexts of their physical environments. While the environmental systems and timescales are dramatically different, continuing to motivate the separation into distinct fields, similarities in the underlying objective present opportunities to find efficiencies, reduce overlap, and minimize duplication of effort. Space Biology and Astrobiology share a common need to understand microbial physiology in extreme environments – whether the ‘built’ spaceflight environment or the natural environments in which many astrobiology studies are conducted. In particular, open questions in each discipline require the development of quantitative frameworks, applicable at the ecosystem level, that support predictive capabilities for environments where observations are sparse. Additionally, both disciplines have a need to prepare, detect, and analyze the (potential) biological signal in complex samples-often in a completely autonomous fashion. The next decade will see NASA Space Biology moving to understand and describe the effects of the beyond low-earth orbit (BLEO) spaceflight environment on living systems. This new direction will dramatically reduce the opportunities for ground-based analysis of space-flown samples, driving space biology investigations towards fully autonomous experiments and missions. At the same time, astrobiology life detection missions aimed at detecting biosignatures on Mars and icy moons in the outer solar system could benefit from fully automated sample processing and analysis. There are opportunities to leverage instrument and method development between both disciplines within the context of these BLEO missions.

Astrobiology↗