Search bioRxiv⌕ Search

Biology subjects

Siboni, N.

Publications and source records attributed to Siboni, N..

2 recordsLinked to original sources

A catastrophic marine mortality event caused by a complex algal bloom including the novel brevetoxin producer, Karenia cristata (Dinophyceae)

Harmful algal blooms of Karenia brevis (Dinophyceae) are a global anomaly, occurring in one location worldwide, causing severe marine and acute human impacts via brevetoxins (BTX). During 2025 an unprecedented, currently ongoing, mass marine mortality occurred in South Australia, across an area of [~]20,000 km2, persisting for >6 months, resulting in the deaths of [~]106 marine animals of >550 taxa, with human health impacts. Using custom metabarcoding, long-read sequencing and targeted quantitative PCR, we characterized the microalgal assemblage. Karenia cristata dominated over the sampling area, in an assemblage with four other Karenia species with varied abundances spatially and temporally. High abundances of K. cristata appeared in the austral autumn, and hydrodynamic processes appear to have entrained cells coastward in the semi-enclosed seas. We isolated the species and characterized it using light and electron microscopy, liquid chromatography mass spectrometry and toxicity assays. We show for the first time that the rare and little known K. cristata produces significant BTX with a profile (BTX-2, -3, -B5), differing from K. brevis, with toxicological effects. These findings reveal a novel, significant BTX-producing Karenia, which considering its substantial detrimental marine ecosystem impacts, is an emerging international threat with unknown consequences in changing ocean conditions.

ecology↗

Offline RL for generative design of protein binders

Offline Reinforcement Learning (RL) offers a compelling avenue for solving RL problems without the need for interactions with an environment, which may be expensive or unsafe. While online RL methods have found success in various domains, such as de novo Structure-Based Drug Discovery (SBDD), they struggle when it comes to optimizing essential properties derived from protein-ligand docking. The high computational cost associated with the docking process makes it impractical for online RL, which typically requires hundreds of thousands of interactions during learning. In this study, we propose the application of offline RL to address the bottleneck posed by the docking process, leveraging RLs capability to optimize non-differentiable properties. Our preliminary investigation focuses on using offline RL to conditionally generate drugs with improved docking and chemical properties.

bioinformatics↗