RESUMO
We seek to transform how new and emergent variants of pandemic-causing viruses, specifically SARS-CoV-2, are identified and classified. By adapting large language models (LLMs) for genomic data, we build genome-scale language models (GenSLMs) which can learn the evolutionary landscape of SARS-CoV-2 genomes. By pre-training on over 110 million prokaryotic gene sequences and fine-tuning a SARS-CoV-2-specific model on 1.5 million genomes, we show that GenSLMs can accurately and rapidly identify variants of concern. Thus, to our knowledge, GenSLMs represents one of the first whole genome scale foundation models which can generalize to other prediction tasks. We demonstrate scaling of GenSLMs on GPU-based supercomputers and AI-hardware accelerators utilizing 1.63 Zettaflops in training runs with a sustained performance of 121 PFLOPS in mixed precision and peak of 850 PFLOPS. We present initial scientific insights from examining GenSLMs in tracking evolutionary dynamics of SARS-CoV-2, paving the path to realizing this on large biological data.
RESUMO
BACKGROUND: Childhood asthma has substantial effects on children's health. It is important to identify influencing factors in early life in the development of childhood asthma. We aim to evaluate the effects of early-life factors and indoor environmental exposure on childhood asthma in Chongqing, China. METHOD: We designed a case-control study to enrol children with asthma aged 3 to < 14 years old and controls in Chongqing, China. The "Children's Early Life and Indoor Environment Survey" was used to collect the early-life factors and indoor environmental exposure of foetuses in utero and of infants during the first 3 years of life. A multivariate logistic regression model was used to evaluate the association between independent variables and childhood asthma and the interaction of early-life factors and environmental exposure. RESULTS: A total of 160 asthma cases and 247 controls were included in this study. The mean ages of the cases and controls were 5.53 ± 1.88 and 5.72 ± 2.34 years, respectively (P = 0.192). Early-life factors and indoor environmental exposure were independently associated with childhood asthma. Infectious diseases of the respiratory system in children under 3 years old [adjusted odds ratio (OR) = 5.76, 95% confidence interval (CI) 2.49-13.30], bedroom air conditioner use (adjusted OR = 4.61, 95% CI 1.45-14.64), and bedroom dampness/mould (adjusted OR = 2.98, 95% CI 1.54-5.75) ranked as the three most significant exposures associated with the risk of childhood asthma. Other factors associated with an increased risk of childhood asthma included second-hand smoke exposure in early life (adjusted OR = 1.93, 95% CI 1.24-3.00), neonatal pneumonia (adjusted OR = 1.90, 95% CI 1.05-3.42) and maternal allergic diseases during pregnancy (adjusted OR = 2.13, 95% CI 1.10-4.10). The interaction effects of child second-hand smoke exposure with other covariates were not found to be statistically significant. CONCLUSIONS: Early-life factors and indoor environmental exposure are closely related to childhood asthma in Chongqing, China. Further interventions and management in the early life of children should be considered to prevent and control childhood asthma in Chongqing and similar cities.