3 Hits in 3.8 sec

Modelling and Recognition of Protein Contact Networks by Multiple Kernel Learning and Dissimilarity Representations

Alessio Martino, Enrico De Santis, Alessandro Giuliani, Antonello Rizzi
2020 Entropy  
This equips the system with a two-fold knowledge discovery phase: by analysing the weights, it is possible to check which representations are more suitable for solving the classification problem, whereas  ...  In this paper, we propose an hybrid classification system based on a linear combination of multiple kernels defined over multiple dissimilarity spaces.  ...  The main idea in order to build a model through the One-Class Classifier is to use a clustering-evolutionary hybrid technique [81, 82] .  ... 
doi:10.3390/e22070794 pmid:33286565 pmcid:PMC7517365 fatcat:w66fbfzgjncydlg5fmjcsxffdm

Evolutionary Genomics of Grass Powdery Mildew (Blumeria graminis)

Fabrizio Menardo
Acknowledgments We want to thank Susanne Brunner for advice and support, Prof. Dinoor, Veronique Troch and  ...  The two red lines represent the threshold of two log2 fold change in expression (equal to a 4-fold change).  ...  These phenotypic differences could then be correlated to the genomic clusters.  ... 
doi:10.5167/uzh-131833 fatcat:ybcfxurr3fabphpxmlsbwo3rqm

Persistent Infection and Long-Term Carriage of Typhoidal and Nontyphoidal Salmonellae

Ohad Gal-Mor
2018 Clinical Microbiology Reviews  
Some pathogens can establish transient colonization only, but others are capable of infecting their host for many years or even for a lifetime.  ...  Here, following a general summary of Salmonella pathogenicity, host specificity, epidemiology, and laboratory diagnosis, I review the current knowledge about Salmonella persistence and discuss the relevant  ...  This methodology provides high-resolution and reproducible isolate typing and has been successfully implemented for analyzing disease clusters, evolutionary relationships between isolates, and outbreak  ... 
doi:10.1128/cmr.00088-18 pmid:30487167 fatcat:cbwejnmsxjgg3ajnuubuz2b4by