November 2022 Issue

November issue now live

Han, T., Kather, J.N., Pedersoli, F. et al. Image prediction of disease progression for osteoarthritis by style-based manifold extrapolation.

Nature Machine Intelligence is a Transformative Journal; authors can publish using the traditional publishing route OR via immediate gold Open Access.

Our Open Access option complies with funder and institutional requirements.


  • The lack of generalizability and reproducibility of machine learning models in medical applications is increasingly recognized as a substantial barrier to implementing such approaches in real-world clinical settings. Highlighting this issue, Jie Cao et al. aim to reproduce a recent acute kidney injury prediction model, and find persistent discrepancies in model performance in different subgroups.

    • Jie Cao
    • Xiaosong Zhang
    • Karandeep Singh
  • A promising area for deep learning is in modelling complex physical processes described by partial differential equations (PDEs), which is computationally expensive for conventional approaches. An operator learning approach called DeepONet was recently introduced to tackle PDE-related problems, and in new work, this approach is extended with transfer learning, which transfers knowledge obtained from learning to perform one task to a related but different task.

    • Somdatta Goswami
    • Katiana Kontolati
    • George Em Karniadakis
  • Predicting the properties of a molecule from its structure with high accuracy is a crucial problem in digital drug design. Instead of sequence features, Zeng and colleagues use an image representation of a large collection of bioactive molecules to pretrain a model that can be fine-tuned on specific property prediction tasks.

    • Xiangxiang Zeng
    • Hongxin Xiang
    • Feixiong Cheng
  • Physical dynamical processes can be modelled with differential equations that may be solved with numerical approaches, but this is computationally costly as the processes grow in complexity. In a new approach, dynamical processes are modelled with closed-form continuous-depth artificial neural networks. Improved efficiency in training and inference is demonstrated on various sequence modelling tasks including human action recognition and steering in autonomous driving.

    • Ramin Hasani
    • Mathias Lechner
    • Daniela Rus
    Article Open Access
  • Computational methods are important for interpreting missense variants in genetic studies and clinical testing. Zhang and colleagues develop a method based on graph attention neural networks to predict pathogenic missense variants. The method pools information from functionally correlated positions and can improve the interpretation of missense variants.

    • Haicang Zhang
    • Michelle S. Xu
    • Yufeng Shen
  • AI promises to bring many benefits to healthcare and research, but mistrust has built up owing to many instances of harm to under-represented communities. To amend this, participatory approaches can directly involve communities in AI research that will impact them. An important element of such approaches is ensuring that communities can take control over their own data and how they are shared.

  • The use of decision-support systems based on artificial intelligence approaches in antimicrobial prescribing raises important moral questions. Adopting ethical frameworks alongside such systems can aid the consideration of infection-specific complexities and support moral decision-making to tackle antimicrobial resistance.

    • William J. Bolton
    • Cosmin Badea
    • Timothy M. Rawson
  • Indigenous peoples are under-represented in genomic datasets, which can lead to limited accuracy and utility of machine learning models in precision health. While open data sharing undermines rights of Indigenous communities to govern data decisions, federated learning may facilitate secure and community-consented data sharing.

    • Nima Boscarino
    • Reed A. Cartwright
    • Krystal S. Tsosie
  • We introduced reusability reports, an article type to highlight code reusability, almost two years ago. On the basis of the results and positive feedback from authors and referees, we remain enthusiastic about the format.

  • To deliver value in healthcare, artificial intelligence and machine learning models must be integrated not only into technology platforms but also into local human and organizational ecosystems and workflows. To realize the promised benefits of applying these models at scale, a roadmap of the challenges and potential solutions to sociotechnical transferability is needed.

    • Batia Mishan Wiesenfeld
    • Yin Aphinyanaphongs
    • Oded Nov