In recent years, the frequency and intensity of wildfires have increased due to the effects of climate change. To date, many established wildfire policies have been ineffective, and alternative approaches are therefore needed in order to reduce the damage caused by such wildfire events. In this issue, Hussam Mahmoud suggests that future wildfire models should take inspiration from epidemic network modeling to predict the damage to individual buildings, as well as to improve our understanding of the impact of different mitigation strategies on the community as a whole.

See Mahmoud

Recently, there has been growing interest and enthusiasm in using digital twins to accelerate scientific discovery and to help researchers and stakeholders with critical decision-making tasks. While the industrial and engineering spaces have seen more developments in digital twin technology, multiple other areas of science — from climate sciences to medical and social sciences — have also realized the potential of digital twins for bringing value and innovation to myriad applications. Nevertheless, many challenges still need to be addressed before the research community can bring the promise of digital twins to fruition. This issue presents a Focus in which we highlight the state of the art, challenges and opportunities in the development and use of digital twins across different domains, with the goal of fostering discussion and collaboration within the computational science community regarding this burgeoning field.

See Focus and Editorial

Data-driven discovery of algorithms is an important task for uncovering the underlying logic and rules behind experimental data and can be potentially used by researchers for generating new insights hidden in high-dimensional data. In this issue, Milo M. Lin et al. introduce an approach that makes use of a neurobiologically inspired deep learning algorithm for writing interpretable and executable computer code from data. The method is able to discover algorithms that perform very similarly to or that outperform human-designed ones. The cover image depicts source code that was transformed into an image featuring bands and gaps, similar to a DNA autoradiogram.

See Milo M. Lin et al. and Joseph Bakarji

While the socio-demographic factors that play an important role in human lives are well understood, accurately predicting life outcomes has not been possible. In this issue, Sune Lehmann et al. introduce a machine learning approach, based on language processing techniques, that can predict different aspects of human lives. The proposed model — called ‘life2vec’ — establishes relationships between concepts, captured by an embedding space, that form the foundation for the predictions of life outcomes. The image depicts such an embedding space as it converges, where the white dots represent individuals and white lines represent how they move as the model is optimized. The shades of blue represent the density of points: the brighter the blue, the higher the density.

See Sune Lehmann et al.