North China University of Science & Technology

Major depressive disorder (MDD) impacts >300 million individuals worldwide, highlighting a significant public health issue. However, the uneven distribution of medical resources and the complexity of diagnostic methods have resulted in inadequate attention to this disorder in numerous countries and regions.

This paper introduces a high-performance MDD diagnosis tool named MDD-LLM, an AI-driven framework that utilizes fine-tuned large language models (LLMs) and extensive real-world samples to tackle challenges in MDD diagnosis. Specifically, we select 274,348 individual records from the UK Biobank cohort and design three tabular data transformation methods to create a large corpus for training and evaluating the proposed method. To illustrate the advantages of MDD-LLM, we perform comprehensive experiments and provide several comparative analyses against existing model-based solutions across multiple evaluation metrics.

Experimental results show that MDD-LLM (70B) achieves an accuracy of 0.8378 and an AUC of 0.8919 (95 % CI: 0.8799-0.9040), significantly outperforming existing machine and deep learning frameworks for MDD diagnosis. Given the limited exploration of LLMs in MDD diagnosis, we examine numerous factors that may influence the performance of our proposed method, including tabular data transformation techniques and different fine-tuning strategies. Furthermore, we also analyze the model's interpretability, requiring the MDD-LLM to explain its predictions and provide corresponding reasons.

This paper investigates the application of LLMs and large-scale training samples for diagnosing MDD. The findings indicate that LLMs-driven schemes offer significant potential for accuracy, robustness, and interpretability in MDD diagnosis compared to traditional model-based solutions.

Polymer prodrug nanoparticles have become an emerging drug delivery system in cancer therapy due to their high drug loading. However, their poor drug release and lack of tumor cell targeting limit their clinical application.

This study aimed to prepare targeted and reduction-reactive polyprodrug nanocarriers based on curcumin (CUR) for co-delivery of doxorubicin (DOX), labeled as DOX/HAPCS NPs, and to investigate their anticancer activity.

The polymer was synthesized and characterized by chemical method. The drug loading and drug release behavior of DOX and CUR in polymer nanoparticles were determined. Moreover, the antitumor effects of polymer nanoparticles were evaluated using an MTT experiment and tumor inhibition experiment, and the synergistic effect of co-delivered DOX and CUR was explored.

The particle size of DOX/HAPCS NPs was 152.5nm, and the potential was about -26.74 mV. The drug-carrying capacity of DOX and CUR was about 7.56% and 34.75%, respectively, indicating high drug-carrying capacity and good stability. DOX and CUR released over 90% within 24 hours in the tumor environment. Compared with free DOX, DOX/HAPCS NPs demonstrated significantly enhanced cell and tumor inhibitory effects (P<0.05) in vivo and in vitro and changed drug distribution to avoid toxic side effects on normal tissues. The combined index showed that DOX and CUR showed synergistic anticancer effects at a set ratio.

The prepared reduction-responsive targeted polymer nanomedical DOX/HAPCS NPs exhibited a synergistic anti-cancer effect, with high drug loading capacity and the ability to release drugs in proportion, making it a promising polymer nanoparticle drug delivery system.

A novel biomass thermo-responsive composite gel (KGM@SSG) synthesized from konjac glucomannan (KGM), steel slag (SS), and silica fume (SF) prevents spontaneous coal combustion.This gel shows commendable flowability, accelerated gelation, excellent moisture retention, and superior flame-retardant properties.Compositional and structural changes were analyzed using XRD, FTIR, SEM, and TG.Results reveal that the amorphous compound Calcium (Aluminum) Silicate Hydrate (C-(A)-S-H), derived from hydrated SS and SF, integrates into KGM′s three-dimensional network, enhancing robustness.Thermogravimetric anal., thermal experiments, and in-situ IR spectroscopy confirm that KGM@SSG mitigates temperature surges during ignition and inhibits coal decompositionWhen applied to coal, it moisturizes the surface, forms a protective coating, slows temperature increases, blocks oxygen adsorption, reduces interactions between active components and oxygen species, and prevents thermal accumulation, supported by mol. dynamics simulations.