The geographic ranges in which species live is a function of many factors underlying ecological and evolutionary contingencies. Observing the geographic range of an individual species provides valuable information about these historical contingencies for a lineage, determining the distribution of many distantly related species in tandem provides information about large-scale constraints on evolutionary and ecological processes generally. We present a linear regression method that allows for the discrimination of various hypothetical biogeographical models for determining which landscape distributional pattern best matches data from the fossil record. The linear regression models used in the discrimination rely on geodesic distances between sampling sites (typically geologic formations) as the independent variable and three possible dependent variables: Dice/Sorensen similarity; Euclidean distance; and phylogenetic community dissimilarity. Both the similarity and distance measures are useful for full-community analyses without evolutionary information, whereas the phylogenetic community dissimilarity requires phylogenetic data. Importantly, the discrimination method uses linear regression residual error to provide relative measures of support for each biogeographical model tested, not absolute answers or p-values. When applied to a recently published dataset of Campanian pollen, we find evidence that supports two plant communities separated by a transitional zone of unknown size. A similar case study of ceratopsid dinosaurs using phylogenetic community dissimilarity provided no evidence of a biogeographical pattern, but this case study suffers from a lack of data to accurately discriminate and/or too much temporal mixing. Future research aiming to reconstruct the distribution of organisms across a landscape has a statisticalbased method for determining what biogeographic distributional model best matches the available data.

Polygenic risk score (PRS) is a quantity that aggregates the effects of variants across the genome and estimates an individual's genetic predisposition for a given trait. PRS analysis typically contains two input data sets: base data for effect size estimation and target data for individual-level prediction. Given the availability of large-scale base data, it becomes more common that the ancestral background of base and target data do not perfectly match. In this paper, we treat the GWAS summary information obtained in the base data as knowledge learned from a pre-trained model, and adopt a transfer learning framework to effectively leverage the knowledge learned from the base data that may or may not have similar ancestral background as the target samples to build prediction models for target individuals. Our proposed transfer learning framework consists of two main steps: (1) conducting false negative control (FNC) marginal screening to extract useful knowledge from the base data; and (2) performing joint model training to integrate the knowledge extracted from base data with the target training data for accurate trans-data prediction. This new approach can significantly enhance the computational and statistical efficiency of joint-model training, alleviate over-fitting, and facilitate more accurate trans-data prediction when heterogeneity level between target and base data sets is small or high.