Search for: All records

Some social animals are highly cooperative creatures that live in tight-knit colonies. Bees and ants are perhaps the most well-known examples of social insects, while Damaraland mole-rats and naked mole-rats, two rodent species found in southern and eastern Africa, are among the most cooperative mammal species. In these colony-forming animals, only one or a few females reproduce and these fertile females are frequently referred to as “queens”. When an animal becomes a queen, her body shape can change dramatically to support the demands of high fertility and frequent reproduction. The molecular basis of such changes has been well-described in social insects. However, they are poorly understood in mammals. To address this knowledge gap, Johnston et al. studied how transitioning to queen status affects bone growth and structural integrity in Damaraland mole-rats, as well as body shape and size. The experiments compared non-breeding female mole-rats with other adult females recently paired with a male to become the sole breeder of a new colony. Johnston et al. also used bone-derived cells grown in the laboratory to assess underlying gene regulatory changes in new queen mole-rats. Johnston et al. showed that transitioning to the role of queen leads to a cascade of skeletal changes accompanied by shifts in the regulation of genetic pathways linked to bone growth. Queen mole-rats show accelerated growth in the spinal column of their lower back. These bones are called lumbar vertebrae and this likely allows them to have larger litters. However, queen mole-rats also lose bone growth potential in their leg bones and develop thinner thigh bones, which may increase the risk of bone fracture. Therefore, unlike highly social insects, mole-rats do not seem to have escaped the physical costs of intensive reproduction. This work adds to our understanding of the genes and physical traits that have evolved to support cooperative behaviour in social animals, including differences between insects and mammals. It also shows, with a striking example, how an animal’s genome responds to social cues to produce a diverse range of traits that reflect their designated social role. 

Abstract In this manuscript we develop a deep learning algorithm to improve estimation of rates of progression and prediction of future patterns of visual field loss in glaucoma. A generalized variational auto-encoder (VAE) was trained to learn a low-dimensional representation of standard automated perimetry (SAP) visual fields using 29,161 fields from 3,832 patients. The VAE was trained on a 90% sample of the data, with randomization at the patient level. Using the remaining 10%, rates of progression and predictions were generated, with comparisons to SAP mean deviation (MD) rates and point-wise (PW) regression predictions, respectively. The longitudinal rate of change through the VAE latent space (e.g., with eight dimensions) detected a significantly higher proportion of progression than MD at two (25% vs. 9%) and four (35% vs 15%) years from baseline. Early on, VAE improved prediction over PW, with significantly smaller mean absolute error in predicting the 4 th , 6 th and 8 th visits from the first three (e.g., visit eight: VAE8: 5.14 dB vs. PW: 8.07 dB; P < 0.001). A deep VAE can be used for assessing both rates and trajectories of progression in glaucoma, with the additional benefit of being a generative technique capable of predicting future patterns of visual field damage.