Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets.

Accurate and comprehensive extraction of information from high-dimensional single cell datasets necessitates faithful visualizations to assess biological populations. A state-of-the-art algorithm for non-linear dimension reduction, t-SNE, requires multiple heuristics and fails to produce clear representations of datasets when millions of cells are projected. We develop opt-SNE, an automated toolkit for t-SNE parameter selection that utilizes Kullback-Leibler divergence evaluation in real time to tailor the early exaggeration and overall number of gradient descent iterations in a dataset-specific manner. The precise calibration of early exaggeration together with opt-SNE adjustment of gradient descent learning rate dramatically improves computation time and enables high-quality visualization of large cytometry and transcriptomics datasets, overcoming limitations of analysis tools with hard-coded parameters that often produce poorly resolved or misleading maps of fluorescent and mass cytometry data. In summary, opt-SNE enables superior data resolution in t-SNE space and thereby more accurate data interpretation.

Symbiotic legume nodules employ both rhizobial exo- and endo-hydrogenases to recycle hydrogen produced by nitrogen fixation.

BACKGROUND: In symbiotic legume nodules, endosymbiotic rhizobia (bacteroids) fix atmospheric N(2), an ATP-dependent catalytic process yielding stoichiometric ammonium and hydrogen gas (H(2)). While in most legume nodules this H(2) is quantitatively evolved, which loss drains metabolic energy, certain bacteroid strains employ uptake hydrogenase activity and thus evolve little or no H(2). Rather, endogenous H(2) is efficiently respired at the expense of O(2), driving oxidative phosphorylation, recouping ATP used for H(2) production, and increasing the efficiency of symbiotic nodule N(2) fixation. In many ensuing investigations since its discovery as a physiological process, bacteroid uptake hydrogenase activity has been presumed a single entity. METHODOLOGY/PRINCIPAL FINDINGS: Azorhizobium caulinodans, the nodule endosymbiont of Sesbania rostrata stems and roots, possesses both orthodox respiratory (exo-)hydrogenase and novel (endo-)hydrogenase activities. These two respiratory hydrogenases are structurally quite distinct and encoded by disparate, unlinked gene-sets. As shown here, in S. rostrata symbiotic nodules, haploid A. caulinodans bacteroids carrying single knockout alleles in either exo- or-endo-hydrogenase structural genes, like the wild-type parent, evolve no detectable H(2) and thus are fully competent for endogenous H(2) recycling. Whereas, nodules formed with A. caulinodans exo-, endo-hydrogenase double-mutants evolve endogenous H(2) quantitatively and thus suffer complete loss of H(2) recycling capability. More generally, from bioinformatic analyses, diazotrophic microaerophiles, including rhizobia, which respire H(2) may carry both exo- and endo-hydrogenase gene-sets. CONCLUSIONS/SIGNIFICANCE: In symbiotic S. rostrata nodules, A. caulinodans bacteroids can use either respiratory hydrogenase to recycle endogenous H(2) produced by N(2) fixation. Thus, H(2) recycling by symbiotic legume nodules may involve multiple respiratory hydrogenases.