Premise: Parasitic plants and their hosts are emerging model systems for studying genetic variation in species interactions across environments. The parasitic plant Striga hermonthica (witchweed) attacks a range of cereal crop hosts in Africa. Striga hermonthica exhibits substantial genetic variation in host preference and in specificity versus generalism. Some of this variation is locally adapted, but the genetic basis of specialization on certain hosts is unknown. Methods: We present an alignment-free analysis of population diversity in S. hermonthica using whole genome sequencing (WGS) data for 68 individuals from western Kenya. We validate our reference-free approach with germination experiments and a de novo assembled draft genome. Results: K-mer based analyses reveal high genome-wide diversity within a single field, similar to values between individuals collected 100 km apart or farther. Analysis of host-associated k-mers implicated genes involved in development of the parasite haustorium (a specialized structure used to establish vascular connections with host roots) and a potential role of chemocyanins in molecular host-parasitic plant interactions. Conversely, no phenotypic or genomic evidence was observed suggesting host-specific selection on parasite response to strigolactones, hormones exuded by host roots and required for parasite germination. Conclusions: This study demonstrates the utility of WGS for plant species with large, complex genomes and no available reference. Contrasting with theory emphasizing the role of early recognition loci for genotype specificity, our findings support host-specific selection on later interaction stages, suggesting recurring host-specific selection each generation alternating with homogenizing gene flow.

Alzheimer′s Disease (AD) is a degenerative brain disease and is the most common cause of dementia. Despite being a common disease, AD is poorly understood. Current medical treatments for AD are aimed at slowing the progression of the disease. So early detection of AD is important to intervene at an early stage of the disease. In recent years, by using machine learning predictive algorithms, assisted clinic diagnosis has received great attention due to its success of machine learning advances in the domains of computer vision. In this study, we have combined brain MRI imaging features and the features of other datatypes, and adopted various models, including XGBoost, logistic regression, and k-Nearest Neighbors, to improve AD diagnosis. We evaluated the models on the benchmark dataset of Alzheimer′s Disease Neuroimaging Initiative. Experiment results show that the logistic regression model is the top performer in terms of evaluation metrics of precision, recall, and F1-score. The prediction of the models could provide valuable information for diagnosis and prognosis of patients with suspected Alzheimer′s disease. The XGBoost model achieves a comparable performance and has the potential to serve as a valuable diagnostic tool for patients with suspected AD with its self-validation by re-discovering previously known associations with AD.