We demonstrate the effectiveness of the algorithms on three gene expression datasets for cancer, showing they compare well with alternative kernel-based techniques. By automatically incorporating feature selection, accurate classifiers can be constructed utilizing very few features and with minimal hand-tuning. We argue that the feature selection is meaningful and some of the highlighted genes appear to be medically important.
Clubroot disease caused by Plasmodiophora brassicae can lead to serious yield losses in crucifers such as Brassica napus. In this study, 323 bacterial strains were isolated from the rhizosphere of severely diseased B. napus in Dangyang county, Hubei province, China. Antagonistic strains were first identified based on dual culture inhibition zones with Fusarium oxysporum and Magnaporthe oryzae. These were then further screened in germination inhibition and viability assays of resting spores of P. brassicae. Finally, eight of the antagonistic strains were found to significantly reduce the disease severity of clubroot by more than 40% under greenhouse conditions, and two strains, F85 and T113, were found to have efficacy of more than 80%. Root hair infection experiments showed that F85 and T113 can inhibit early infection of root hairs, reduce the differentiation of primary plasmodia of P. brassicae, and inhibit formation of secondary zoosporangia. Based on sequence analysis of 16S rDNA gene, gyrA gene and 22 housekeeping genes as well as carbon source utilization analysis, the F85 was identified as Bacillus velezensis and T113 as Bacillus amyloliquefaciens. Genome analysis, PCR and RT-PCR detection revealed that both F85 and T113 harbor various antibiotic biosynthesis gene clusters required to form peptides with antimicrobial activity. To our knowledge, this is the first report of B. velezensis as a biocontrol agent against clubroot disease.
Database-driven Cognitive Radio Network (CRN) has been proposed to replace the requirement of spectrum sensing of terminal devices so that the operation of users is simplified. However, location privacy issues introduce a big challenge for securing database-driven CRN due to spectrum availability information. The existing works consider either PU or SU's location privacy while not the both. In this study, we identify a unified attack framework in which a curious user could infer a target's location based on the spectrum availability/utilization information. Further, we propose a location privacy protection mechanism, which allows both SU and PU to protect their location privacy by adopting a series of countermeasures. The location privacy and spectrum utility are the trade-off. In the countermeasures of location privacy preserving spectrum query process, both SU and database aim to maximize the location privacy with constraints of spectrum utility. Thus, they can obtain higher location privacy level with sacrifice of spectrum utility as long as the spectrum utility meets the requirements. We evaluate the unified attack and defence approaches based on simulation and demonstrate the effectiveness of the proposed location privacy preserving approaches.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.