Max Hung Nguyen was 17 when he began looking at a problem that astronomers have been studying for decades: whether the chemical make-up of a star can offer clues about the planets that form around it. The Leland High School student from San Jose used data from Nasa’s Exoplanet Archive and the Hypatia Catalogue for his Regeneron Science Talent Search project, Star2Planet: Decoding and Predicting the Stellar Signatures of Planet Formation. Instead of relying only on iron to represent a star’s metallicity, Nguyen included carbon, oxygen, magnesium and silicon in his analysis. His results indicated that stars hosting giant planets tend to be richer in heavy elements, while rocky planets occur around a broader range of stars. He then trained an AI model to estimate missing elemental data, opening another route for identifying stars that could be useful targets in the search for planets.
How Max Hung Nguyen used Nasa data and AI to predict stars likely to host giant planets
In astronomy, metallicity does not simply mean that a star contains metal in the everyday sense. It refers to the abundance of elements heavier than hydrogen and helium. Those elements matter because the material from which planets eventually develop is tied to the chemical make-up of the system surrounding a young star.Nguyen’s project focused on whether that relationship looks different depending on the kind of planet being formed. According to the Society for Science, his analysis examined planetary-system metallicity using information from the Nasa Exoplanet Archive and the Hypatia Catalogue. The Nasa archive brings together data on exoplanets and their host stars, including stellar and planetary properties, while Hypatia is a large compilation of measured elemental abundances in stars.
Why stellar metallicity matters for predicting giant and rocky planets
A common way of describing stellar metallicity is through iron abundance. It is a useful measurement, but Nguyen’s approach asked whether a broader chemical picture could do a better job. His version incorporated carbon, oxygen, magnesium and silicon alongside iron when assessing the composition of planetary systems.That change mattered to his analysis. Those additional elements improved the predictions of planet formation. The distinction was particularly interesting when the planets were separated by type. Systems containing gas giants tended to be associated with stars richer in heavy elements, whereas rocky planets appeared across a much wider range of stellar compositions. The result did not suggest that one particular chemical profile guarantees an Earth-like planet. Instead, it pointed to a more complicated relationship between a star’s ingredients and the worlds that develop around it.
How AI helped predict the chemical composition of stars
There was a second problem to tackle. Even when astronomers know a star’s basic properties, detailed measurements of its elemental abundances are not necessarily available. That creates a gap if those measurements are useful for predicting the kinds of planets a system might contain.Nguyen used AI as a way of addressing that gap. As described by the Society for Science, he trained a model to estimate the heavy-element content of stars when those observations were missing. The idea was to use information that is more readily available about a star to make an informed estimate of its chemical composition, and then connect that estimate with the likelihood of different planet types. It turns the project from a simple comparison of existing observations into a possible screening method for stars that have not yet been studied in the same level of detail.
How Nasa’s Exoplanet Archive and Hypatia Catalogue support Nguyen’s research
The datasets behind the project are substantial. Nasa describes its Exoplanet Archive as an online astronomical catalogue and data service that brings together information on exoplanets and their host stars, with tools for searching and analysing the data. The archive also includes data from major planet-hunting programmes and currently lists thousands of confirmed planets and candidates. Hypatia, meanwhile, brings together stellar abundance measurements from numerous literature sources, allowing researchers to examine the chemical composition of stars across multiple elements.For Nguyen, the value of combining these sources was not simply in producing another catalogue of known planets. His model points towards a way of using stellar chemistry as an early clue when deciding which systems might deserve closer attention. If the composition of a star can be estimated reliably enough, it could help astronomers narrow down which systems are more likely to contain giant planets and which are less strongly associated with them. The project therefore links something happening at stellar scale, the distribution of elements, with the much more distant question of what worlds may be orbiting that star.
Max Hung Nguyen’s work beyond Star2Planet and astronomy
Nguyen’s work on Star2Planet sits alongside activities closer to home. The Society for Science says he attends Leland High School in San Jose, where he co-leads a peer-tutoring programme. He also volunteers with The Cross Team, an organisation that serves people without housing.His interests also extend beyond astronomy and data analysis. Nguyen founded WellnessPlay, a student-run youth organisation that uses gaming as part of its focus on mental health. According to his Society for Science profile, the organisation has grown to 12 chapters and more than 200 members worldwide. He is also a fan of classic rock and learned to play guitar, with the science behind electric guitars becoming another way for him to connect technical ideas with music.The combination gives some context to the project without changing what it set out to do. Star2Planet was a data-driven attempt to examine whether a richer picture of stellar chemistry can tell astronomers something about planetary formation, and whether machine learning can help when that chemical information is incomplete.
