Search for: All records

Abstract The mechanism of mortality plays a large role in how microorganisms in the open ocean contribute to global energy and nutrient cycling. Salps are ubiquitous pelagic tunicates that are a well-known mortality source for large phototrophic microorganisms in coastal and high-latitude systems, but their impact on the immense populations of smaller prokaryotes in the tropical and subtropical open ocean gyres is not well quantified. We used robustly quantitative techniques to measure salp clearance and enrichment of specific microbial functional groups in the North Pacific Subtropical Gyre, one of the largest ecosystems on Earth. We discovered that salps are a previously unknown predator of the globally abundant nitrogen fixer Crocosphaera; thus, salps restrain new nitrogen delivery to the marine ecosystem. We show that the ocean's two numerically dominant cells, Prochlorococcus and SAR11, are not consumed by salps, which offers a new explanation for the dominance of small cells in open ocean systems. We also identified a double bonus for Prochlorococcus, wherein it not only escapes salp predation but the salps also remove one of its major mixotrophic predators, the prymnesiophyte Chrysochromulina. When we modeled the interaction between salp mesh and particles, we found that cell size alone could not account for these prey selection patterns. Instead, the results suggest that alternative mechanisms, such as surface property, shape, nutritional quality, or even prey behavior, determine which microbial cells are consumed by salps. Together, these results identify salps as a major factor in shaping the structure, function, and ecology of open ocean microbial communities. 

Abstract Summary Chemsearch is a cross-platform server application for developing and managing a chemical compound library and associated data files, with an interface for browsing and search that allows for easy navigation to a compound of interest, similar compounds or compounds that have desired structural properties. With provisions for access control and centralized document and data storage, Chemsearch supports collaboration by distributed teams. Availability and implementation Chemsearch is a free and open-source Flask web application that can be linked to a Google Workspace account. Source code is available at https://github.com/gem-net/chemsearch (GPLv3 license). A Docker image allowing rapid deployment is available at https://hub.docker.com/r/cgemcci/chemsearch. 

Prediction of side chain conformations of amino acids in proteins (also termed “packing”) is an important and challenging part of protein structure prediction with many interesting applications in protein design. A variety of methods for packing have been developed but more accurate ones are still needed. Machine learning (ML) methods have recently become a powerful tool for solving various problems in diverse areas of science, including structural biology. In this study, we evaluate the potential of deep neural networks (DNNs) for prediction of amino acid side chain conformations. We formulate the problem as image-to-image transformation and train a U-net style DNN to solve the problem. We show that our method outperforms other physics-based methods by a significant margin: reconstruction RMSDs for most amino acids are about 20% smaller compared to SCWRL4 and Rosetta Packer with RMSDs for bulky hydrophobic amino acids Phe, Tyr, and Trp being up to 50% smaller.