Aoki, S. — Page 2 | bioRxiv Science

bioRxiv2020

Massive rhizobial genomic variations associated with partner quality in Lotus-Mesorhizobium symbiosis

In diverse mutualistic relationships, genetic variations in impact on the growth of interacting partners--variations in partner quality--are common, despite the theoretical prediction that selection favoring high-quality partners should eliminate such variations. Here, we investigated how variations in partner quality could be maintained in the nitrogen-fixing mutualism between Lotus japonicus and Mesorhizobium bacteria. We reconstructed de novo assembled full-genome sequences from nine rhizobial symbionts, finding massive variations in the core genome and the contrastingly similar symbiotic islands, indicating recent horizontal gene transfer (HGT) of the symbiosis islands into diverse Mesorhizobium lineages. A cross-inoculation experiment using nine sequenced rhizobial symbionts and 15 L. japonicus accessions revealed extensive quality variations represented by plant growth phenotypes, including genotype-by-genotype interactions. Quality variations were not associated with the presence/absence variations of known symbiosis-related genes in the symbiosis island, but rather, showed significant correlations with the core genome variations, supported by SNP- and kinship matrix-based association analyses. These findings highlight the novel role of HGT of symbiosis islands, which indirectly supply mutations of core genomes into L. japonicus-associated bacteria, thereby contributing to the maintenance of variations in partner quality.

evolutionary biology↗

bioRxiv2017

Sparse ordinal logistic regression and its application to brain decoding

Brain decoding with multivariate classification and regression has provided a powerful framework for characterizing information encoded in population neural activity. Classification and regression models are respectively used to predict discrete and continuous variables of interest. However, cognitive and behavioral parameters that we wish to decode are often ordinal variables whose values are discrete but ordered, such as subjective ratings. To date, there is no established method of predicting ordinal variables in brain decoding. In this study, we present a new algorithm, sparse ordinal logistic regression (SOLR), that combines ordinal logistic regression with Bayesian sparse weight estimation. We found that, in both simulation and analyses using real functional magnetic resonance imaging data, SOLR outperformed ordinal logistic regression with non-sparse regularization, indicating that sparseness leads to better decoding performance. SOLR also outperformed classification and linear regression models with the same type of sparseness, indicating the advantage of the modeling tailored to ordinal outputs. Our results suggest that SOLR provides a principled and effective method of decoding ordinal variables.

neuroscience↗