When Microplastic Is Not Plastic: The Ingestion of Artificial Cellulose Fibers by Macrofauna Living in Seagrass Macrophytodetritus.

Dead leaves of the Neptune grass, Posidonia oceanica (L.) Delile, in the Mediterranean coastal zone, are colonized by an abundant "detritivorous" invertebrate community that is heavily predated by fishes. This community was sampled in August 2011, November 2011, and March 2012 at two different sites in the Calvi Bay (Corsica). Ingested artificial fibers (AFs) of various sizes and colors were found in 27.6% of the digestive tracts of the nine dominant species regardless of their trophic level or taxon. No seasonal, spatial, size, or species-specific significant differences were revealed; suggesting that invertebrates ingest AFs at constant rates. Results showed that, in the gut contents of invertebrates, varying by trophic level, and across trophic levels, the overall ingestion of AFs was low (approximately 1 fiber per organism). Raman spectroscopy revealed that the ingested AFs were composed of viscose, an artificial, cellulose-based polymer. Most of these AFs also appeared to have been colored by industrial dyes. Two dyes were identified: Direct Blue 22 and Direct Red 28. The latter is known for being carcinogenic for vertebrates, potentially causing environmental problems for the P. oceanica litter community. Techniques such as Raman spectroscopy are necessary to investigate the particles composition, instead of relying on fragment size or color to identify the particles ingested by animals.

BioLORD-2023: semantic textual representations fusing large language models and clinical knowledge graph insights.

OBJECTIVE: In this study, we investigate the potential of large language models (LLMs) to complement biomedical knowledge graphs in the training of semantic models for the biomedical and clinical domains. MATERIALS AND METHODS: Drawing on the wealth of the Unified Medical Language System knowledge graph and harnessing cutting-edge LLMs, we propose a new state-of-the-art approach for obtaining high-fidelity representations of biomedical concepts and sentences, consisting of 3 steps: an improved contrastive learning phase, a novel self-distillation phase, and a weight averaging phase. RESULTS: Through rigorous evaluations of diverse downstream tasks, we demonstrate consistent and substantial improvements over the previous state of the art for semantic textual similarity (STS), biomedical concept representation (BCR), and clinically named entity linking, across 15+ datasets. Besides our new state-of-the-art biomedical model for English, we also distill and release a multilingual model compatible with 50+ languages and finetuned on 7 European languages. DISCUSSION: Many clinical pipelines can benefit from our latest models. Our new multilingual model enables a range of languages to benefit from our advancements in biomedical semantic representation learning, opening a new avenue for bioinformatics researchers around the world. As a result, we hope to see BioLORD-2023 becoming a precious tool for future biomedical applications. CONCLUSION: In this article, we introduced BioLORD-2023, a state-of-the-art model for STS and BCR designed for the clinical domain.