Abstract
To understand the prenatal origin of developmental and psychiatric disorders, studies in laboratory animals are imperative. However, the developmental pace differs between humans and animals; hence, corresponding human ages must be estimated to infer the most vulnerable developmental timings in humans. Because rats and mice are extensively used as models in developmental research, a correspondence between human foetal ages and rodents' ages must be precisely determined; thus, developing a translational model is of utmost importance. Optimizing a translational model involves classifying the brain regions according to developmental paces, but previous studies have conducted this classification arbitrarily. Here we used a clustering method and showed that the brain regions can be classified into two groups. To quantify the developmental pace, we gathered data for a range of development events in humans and rodents and created a linear mixed model that translates human developmental timings into the corresponding rat timings. We conducted an automatic classification of brain regions using an EM algorithm and obtained a model to translate human foetal age to rat age. Our model could predict rat developmental timings within 2.5 days of root mean squared error. This result provides useful information for designing animal studies and clinical tests.