We are looking for a postdoctoral candidate to work on developing novel machine-learning approaches to model coevolution across biological scales. The successful candidate will join the Model-Based Evolutionary Genomics Unit (http://bit.ly/mod2evol) at OIST, a small, but prolific University with a unique interdisciplinary atmosphere (http://bit.ly/oist1st), and have the opportunity to develop their own project or join ongoing ones that: - Continue and extend work on using Machine Learning to predict ancestral phenotypes based on ancestral gene content; - Use protein language models, e.g. ESM models, to explore fitness landscapes with the goal of being able to reconstruct context-aware ancestral sequences; - Design and train transformer models trained on extant gene content using a masked language modeling objective to produce robust and high- fidelity context-aware ancestral phenotype predictions and improve ancestral gene content reconstruction. The position offers the freedom to pursue your own research ideas and provides a platform for innovation and discovery in a unit that thrives on intellectual curiosity and interdisciplinary collaboration. The position includes funding to travel to international conferences as well as extended research visits to international collaborators, including Tom Williams at the University of Bristol, Anja Spang at NIOZ in the Netherlands, Phil Hugenholtz at the University of Queensland, and others, provided by the budget of the Model-based Evolutionary Genomics Unit. The position is for two years with a possibility of extension. To apply email gergely-szollosi@oist.jp with a CV including publications and three references, as well as a 1-2 page summary of your past research and your future plans by May 10. 2024. Position open until filled. Context: The Model-Based Evolutionary Genomics Unit's Research Objectives Modelling the dependency between different levels of organisation is crucial for understanding the principles that govern living systems and central to understanding the evolution of Life on Earth, as the structure of biological systems today, from molecules to ecosystems, reflects Life’s ascent on a “co-evolutionary ladder” of intertwined information-environment transitions that began 4 billion years ago (Lenton 2004). In recent decades the broad outlines of the major transitions on this co-evolutionary ladder have started to come into focus, from the early emergence of life to the rise of oxygen, eukaryotes and multiple transitions to multicellularity, but we lack an empirical framework within which to test hypotheses about how, when and why these major transitions in evolution have occurred based on the geochemical and fossil records and the wealth of genomic data available. The Model-Based Evolutionary Genomics develops data-driven models that take into account dependencies between different levels of organisation, which are currently neglected due to computational and modelling constraints. Going beyond model development we apply these methods to major outstanding questions across biological scales in collaboration with a multidisciplinary group of key interdisciplinary researchers. The Okinawa Institute of Science and Technology The Okinawa Institute of Science and Technology Graduate University (OIST; see www.oist.jp) is a dynamic new graduate university of science and technology in Okinawa Prefecture, Japan. The university is located on 85 hectares of protected forestland overlooking beautiful shoreline and coral reefs. The campus is striking architecturally, and the facilities are outstanding (OIST campus video tour: https://youtu.be/OLeylXbZDpo ). There are no academic departments, which facilitates multidisciplinary research. Outstanding resources and equipment are provided and managed to encourage easy access and collaboration. English is the official language of the University, and the university research community is fully international, with more than 50 countries represented. OIST is rapidly gaining recognition in the worldwide academic community as a model for excellence in education and research. OIST offers relocation support, on-campus child care, and a chance to be a part of a diverse international community (bit.ly/oistint) on a unique subtropical island (https://okinawanaturephotography.com/)! For more information please email: gergely-szollosi@oist.jp Gergely J. Szöllősi Transitional Associate Professor Model-Based Evolutionary Genomics Unit モデルベース進化ゲノミクスユニット https://www.oist.jp/research/research-units/modevolgenom Okinawa Institute of Science and Technology Gergely Szöllősi (to subscribe/unsubscribe the EvolDir send mail to golding@mcmaster.ca)