Search

Objective Most existing fine-tuned biomedical large language models (LLMs) focus on enhancing performance in monolingual biomedical question answering and conversation tasks. To investigate the effectiveness of the fine-tuned LLMs on diverse biomedical natural language processing (NLP) tasks in different languages, we present Taiyi, a bilingual fine-tuned LLM for diverse biomedical NLP tasks. Materials and Methods We first curated a comprehensive collection of 140 existing biomedical text mining datasets (102 English and 38 Chinese datasets) across over 10 task types. Subsequently, these corpora were converted to the instruction data used to fine-tune the general LLM. During the supervised fine-tuning phase, a 2-stage strategy is proposed to optimize the model performance across various tasks. Results Experimental results on 13 test sets, which include named entity recognition, relation extraction, text classification, and question answering tasks, demonstrate that Taiyi achieves superior performance compared to general LLMs. The case study involving additional biomedical NLP tasks further shows Taiyi's considerable potential for bilingual biomedical multitasking. Conclusion Leveraging rich high-quality biomedical corpora and developing effective fine-tuning strategies can significantly improve the performance of LLMs within the biomedical domain. Taiyi shows the bilingual multitasking capability through supervised fine-tuning. However, those tasks such as information extraction that are not generation tasks in nature remain challenging for LLM-based generative approaches, and they still underperform the conventional discriminative approaches using smaller language models. [ABSTRACT FROM AUTHOR]

This retrospective study aimed to compare the difference of the levels of white blood cells (WBC), C-reactive protein (CRP), procalcitonin, and D-Dimer in the bloodstream infection (BSI) patients, and their values in distinguishing bacterial categories. A total of 847 bloodstream infection patients were analyzed and divided into Gram-positive BSI (GP-BSI) and Gram-negative BSI (GN-BSI) groups. Most frequently isolated pathogens in GP-BSI were Staphylococcus epidermidis (35.75%), followed by Staphylococcus hominis (18.33%), and Streptococcus haemolyticus (10.16%), while in GN-BSI, Escherichia coli (30.07%), Klebsiella pneumoniae (23.98%), and Acinetobacter baumannii (13.18%) were the most common. The predictive value was evaluated based on three years of patient data, which showed an area under the curve (AUC) of 0.828. It was further validated using two years of data, which yielded an AUC of 0.925. Significant differences existed in the procalcitonin, D-Dimer, and CRP levels between GN-BSI and GP-BSI. The current results provide a more effective strategy for early differential diagnosis in bacterial categorization of BSI when combining WBC, CRP, procalcitonin, and D-Dimer measurements., (© The Author(s) 2024. Published by Oxford University Press on behalf of FEMS.)

The BioRED track at BioCreative VIII calls for a community effort to identify, semantically categorize, and highlight the novelty factor of the relationships between biomedical entities in unstructured text. Relation extraction is crucial for many biomedical natural language processing (NLP) applications, from drug discovery to custom medical solutions. The BioRED track simulates a real-world application of biomedical relationship extraction, and as such, considers multiple biomedical entity types, normalized to their specific corresponding database identifiers, as well as defines relationships between them in the documents. The challenge consisted of two subtasks: (i) in Subtask 1, participants were given the article text and human expert annotated entities, and were asked to extract the relation pairs, identify their semantic type and the novelty factor, and (ii) in Subtask 2, participants were given only the article text, and were asked to build an end-to-end system that could identify and categorize the relationships and their novelty. We received a total of 94 submissions from 14 teams worldwide. The highest F-score performances achieved for the Subtask 1 were: 77.17% for relation pair identification, 58.95% for relation type identification, 59.22% for novelty identification, and 44.55% when evaluating all of the above aspects of the comprehensive relation extraction. The highest F-score performances achieved for the Subtask 2 were: 55.84% for relation pair, 43.03% for relation type, 42.74% for novelty, and 32.75% for comprehensive relation extraction. The entire BioRED track dataset and other challenge materials are available at https://ftp.ncbi.nlm.nih.gov/pub/lu/BC8-BioRED-track/ and https://codalab.lisn.upsaclay.fr/competitions/13377 and https://codalab.lisn.upsaclay.fr/competitions/13378. Database URL: https://ftp.ncbi.nlm.nih.gov/pub/lu/BC8-BioRED-track/https://codalab.lisn.upsaclay.fr/competitions/13377https://codalab.lisn.upsaclay.fr/competitions/13378., (Published by Oxford University Press 2024. This work is written by (a) US Government employee(s) and is in the public domain in the US.)

Background: The interplay between gut microbiota and metabolites in the early stages of sepsis-induced acute kidney injury (SA-AKI) is not yet clearly understood. This study explores the characteristics and interactions of gut microbiota, and blood and urinary metabolites in patients with SA-AKI., Methods: Utilizing a prospective observational approach, we conducted comparative analyses of gut microbiota and metabolites via metabolomics and metagenomics in individuals diagnosed with SA-AKI compared to those without AKI (NCT06197828). Pearson correlations were used to identify associations between microbiota, metabolites, and clinical indicators. The Comprehensive Antibiotic Resistance Database was employed to detect antibiotic resistance genes (ARGs), while Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways informed on metabolic processes and microbial resistance patterns., Results: Our study included analysis of four patients with SA-AKI and five without AKI. Significant disparities in bacterial composition were observed, illustrated by diversity indices (Shannon index: 2.0 ± 0.4 vs. 1.4 ± 0.6, P = 0.230; Simpson index: 0.8 ± 0.1 vs. 0.6 ± 0.2, P = 0.494) between the SA-AKI group and the non-AKI group. N6, N6, N6-Trimethyl-L-lysine was detected in both blood and urine metabolites, and also showed significant correlations with specific gut microbiota (Campylobacter hominis and Bacteroides caccae, R > 0, P < 0.05). Both blood and urine metabolites were enriched in the lysine degradation pathway. We also identified the citrate cycle (TCA cycle) as a KEGG pathway enriched in sets of differentially expressed ARGs in the gut microbiota, which exhibits an association with lysine degradation., Conclusions: Significant differences in gut microbiota and metabolites were observed between the SA-AKI and non-AKI groups, uncovering potential biomarkers and metabolic changes linked to SA-AKI. The lysine degradation pathway may serve as a crucial link connecting gut microbiota and metabolites., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Xu, Xu, Zhu, Mao, Li, Kong, Zhu and Zhang.)

Introduction: The benefits of neurally adjusted ventilatory assist (NAVA) in preterm infants are unclear. This study aimed to explore if noninvasive NAVA is more beneficial for preterm infants than nasal continuous positive airway pressure (NCPAP)., Study Design: Meta-analysis was performed in three clinical trials comprising two randomized controlled trials and one crossover study. We compared NIV-NAVA and NCPAP and reported treatment failure, mortality, and adverse events as the primary outcomes., Results: Three studies including 173 patients (89 of whom underwent NIV-NAVA) were eligible for this meta-analysis. This review found no difference in treatment failure between NIVNAVA and NCPAP (RR 1.09, 95% CI 0.65 to 1.84; RD 0.02, 95% CI -0.10-0.14; I2=33%, P=0.23). Similarly, there was no difference in mortality (RR 1.52, 95% CI 0.51-4.52, heterogeneity not applicable). Compared with NCPAP, NIVNAVA significantly reduced the use of caffeine (RR 0.85, 95% CI 0.74-0.98, I2=71%, P=0.03)., Conclusions: Compared with NCPAP, there is insufficient evidence to conclude on the benefits or harm of NIV-NAVA therapy for preterm infants. The findings of this review should be confirmed using methodologically rigorous and adequately powered clinical trials.