Identifying signs of life on another planet would rank among the top scientific discoveries in history. Two of the most daunting challenges to achieving such a discovery are developing instrumentation to study exoplanet atmospheres and developing a quantitative framework to assess whether life creates distinctive atmospheric biosignatures. Here, David Catling will address the latter challenge, aiming to develop a quantitative framework to identify atmospheric signatures of life on distant planets. While billions have been invested in telescopes capable of studying exoplanet atmospheres, we lack robust methods to interpret this data for signs of life.   The project focuses on assessing how living organisms impact a planet’s atmosphere. As such, the research team makes assumptions about life. Firstly, they assume Earth-familiar microbial life because Earth contains the only known examples of life, and because it’s thought that if life exists elsewhere in the universe, it’s more likely to be microbial rather than plant- or animal-based. Secondly, they assume redox-based metabolism would be universal to any life form because reduction / oxidation (redox) reactions are the only class of chemical reactions that release enough energy to satisfy the high energy demands of organisms. Rather than looking for individual gases that might indicate living systems, Catling proposes examining chemical disequilibrium—multiple gases coexisting that should normally react and eliminate each other—as a more reliable biosignature.   The research team will build an integrated model simulating planetary evolution from lifeless to hosting various biospheres. They'll quantify two potential biosignatures: free energy dissipation (which should increase dramatically with biological activity) and the information content of atmospheric disequilibrium. The final step involves determining which measurable gas abundances and fluxes most reliably indicate biological activity.

This grant will support Professor Zhiyue Lu to develop a theoretical framework to understand biochemical networks that operate far from equilibrium and outside steady states. Living systems exist in far-from-equilibrium states; this characteristic is one of the most striking distinctions between living and non-living systems.   Unlike non-living matter, living organisms demonstrate remarkable sensitivity and complex responses to environmental changes. While most scientific understanding focuses on near-equilibrium or steady-state systems, Lu aims to explore how biochemical networks respond to time-varying environmental conditions.   The research focuses on three key properties of living systems: sensitivity and robustness to environmental changes, ability to manipulate energy to power various processes, and capacity to extract energy from fluctuating environments. Through mathematical modeling and simulation, Lu's team will investigate how network structure influences these properties under different temporal patterns of environmental change. Specifically, they'll study how networks process complex information patterns, combine weak energy sources to power energy-intensive processes (or distribute energy from one source to multiple processes), and extract energy from environments that fluctuate at different timescales.

Scientists believe that before complex cellular life emerged on Earth, simpler “containers” may have served as early "protocells." Droplets containing various chemical systems are one candidate for such cell-like containers, but we currently have a poor understanding of whether such a protocell could have grown and divided—a fundamental requirement for life. Using a technique called electrochemistry, which applies electric current to drive chemical reactions, Professor Julio Alvarez will study how electrical charges affect the behavior of tiny droplets. When droplets become sufficiently charged, they can become unstable and split apart, potentially mimicking cellular division. The research team will use ultra-small electrodes to precisely control chemical reactions that modify droplet charge states. They'll examine various factors including different surfactants (molecules that stabilize droplets), solution environments, and conditions that mimic the crowded, viscous environment inside modern cells. By understanding what controls droplet growth and division, this research could illuminate a critical step in the transition from non-living chemistry to life on early Earth. The findings may also contribute to modern synthetic biology efforts to build artificial cells from scratch.

Discussions around the value of AI in science lack clarity about what ‘AI’ actually entails—from basic neural networks to advanced foundation models—and how these tools are applied in the scientific process. This project will systematically map the use of various AI approaches (machine learning, deep learning, foundation models) across scientific disciplines and research workflows. The team will begin by mining scientific literature for citations of key AI papers. Using large language models and natural language processing, they will assess the degree of AI engagement in citing papers—ranging from simple references for context, to active use in research methods, to modification of the AI models themselves. The study will also explore how adoption varies across disciplines based on model attributes such as size and openness. The focus will remain on AI’s role within the research process itself, rather than on administrative applications. Ultimately, the project will shed light on the evolving role of AI in science and bring greater precision to how we talk about AI’s place in scientific research.

Software Heritage (SWH) is an ambitious and pioneering initiative to archive all software source code ever written. Founded by Roberto Di Cosmo with Stefano Zacchiroli and hosted at INRIA in France, the project systematically collects code from software repositories worldwide. SWH now houses over 347 million software projects comprising nearly 24 billion source files. SWH has become integral to France's open science strategy, which explicitly acknowledges software as a key scientific output. Their tracking dashboard and curation tools provide transparency for software created through French research funding. This project will leverage existing SWH infrastructure to directly support the work of university-based Open Source Program Offices. Following a requirements survey, Di Cosmo has developed a plan to generalize the French dashboard, extend the CodeMeta data standard, and dedicate staff to implementation support. This grant will allow SWH, under the supervision of Morane Gruenpeter, to extend its infrastructure to directly benefit other Sloan grantees. The proposed tools address a critical need for scientific software supporters in universities and could significantly lower barriers to entry for new OSPOs, while also enhancing SWH's capabilities for AI model training.

This initiative aims to enhance the Software Heritage (SWH) archive, the world's largest collection of software source code, by developing advanced compression and search capabilities. This work is particularly timely, given the increasing use of generative AI in software development and maintenance, but also in the use of code to train better-performing generative AI tools. SWH currently stores code using basic compression and search functionalities, which limit users to search only by metadata rather than within the actual code. The project will pursue two parallel research tracks: creating more efficient lossless compression methods and developing novel "code-to-code" search functionality. These improvements would enable researchers to perform sophisticated searches within the SWH codebase and make its computational infrastructure more sustainable and scalable. The enhancements could empower diverse applications. Beyond improving SWH itself, this research has the potential to benefit the broader open-source ecosystem by lowering barriers to using SWH in training next-generation AI models, providing essential tools for verifying provenance of code (possibly AI-generated), and supporting new cybersecurity applications.

As Research Software Engineers (RSEs) become more integrated into academic institutions and scientific grant proposals, there's a growing need to understand their impact, but empirical data on RSEs' technical and scientific contributions remains limited. To address this gap, Carnegie Mellon's James Herbsleb (who studies code contributions and hackathons) and University of Washington's Anissa Tanweer (an expert on data science practices) propose a comprehensive study combining ethnographic research on four university-based RSE teams and their scientific partners, as well as large-scale analysis of networks and code contributions in open-source scientific software projects with RSE involvement Schmidt Sciences will co-fund this initiative, and study sites will include representation from their "Virtual Institute for Scientific Software" program. The research aims to identify challenges, best practices, and impact measurements for RSE teams, with findings targeted toward university administrators and funders who are considering investments in RSE positions—delivering timely insights as this professional role continues to evolve.

Over more than a decade, The Carpentries has become a vital resource for the computational research community, providing critical training and upskilling for researchers through an impressive network of 4,000+ volunteer instructors. In 2024 alone, they trained approximately 7,000 researchers through 278 workshops. Their open curricula extend this impact even further. The Carpentries maintains a core business model based on institutional memberships, supplemented by philanthropy. This grant will primarily fund personnel costs, particularly in curriculum development and community management, plus governance meetings to implement their strategic plan.