Fine-tuning large language models for chemical text mining.

Extracting knowledge from complex and diverse chemical texts is a pivotal task for both experimental and computational chemists. The task is still considered to be extremely challenging due to the complexity of the chemical language and scientific literature. This study explored the power of fine-tuned large language models (LLMs) on five intricate chemical text mining tasks: compound entity recognition, reaction role labelling, metal-organic framework (MOF) synthesis information extraction, nuclear magnetic resonance spectroscopy (NMR) data extraction, and the conversion of reaction paragraphs to action sequences. The fine-tuned LLMs demonstrated impressive performance, significantly reducing the need for repetitive and extensive prompt engineering experiments. For comparison, we guided ChatGPT (GPT-3.5-turbo) and GPT-4 with prompt engineering and fine-tuned GPT-3.5-turbo as well as other open-source LLMs such as Mistral, Llama3, Llama2, T5, and BART. The results showed that the fine-tuned ChatGPT models excelled in all tasks. They achieved exact accuracy levels ranging from 69% to 95% on these tasks with minimal annotated data. They even outperformed those task-adaptive pre-training and fine-tuning models that were based on a significantly larger amount of in-domain data. Notably, fine-tuned Mistral and Llama3 show competitive abilities. Given their versatility, robustness, and low-code capability, leveraging fine-tuned LLMs as flexible and effective toolkits for automated data acquisition could revolutionize chemical knowledge extraction.

Deep representation learning of chemical-induced transcriptional profile for phenotype-based drug discovery.

Artificial intelligence transforms drug discovery, with phenotype-based approaches emerging as a promising alternative to target-based methods, overcoming limitations like lack of well-defined targets. While chemical-induced transcriptional profiles offer a comprehensive view of drug mechanisms, inherent noise often obscures the true signal, hindering their potential for meaningful insights. Here, we highlight the development of TranSiGen, a deep generative model employing self-supervised representation learning. TranSiGen analyzes basal cell gene expression and molecular structures to reconstruct chemical-induced transcriptional profiles with high accuracy. By capturing both cellular and compound information, TranSiGen-derived representations demonstrate efficacy in diverse downstream tasks like ligand-based virtual screening, drug response prediction, and phenotype-based drug repurposing. Notably, in vitro validation of TranSiGen's application in pancreatic cancer drug discovery highlights its potential for identifying effective compounds. We envisage that integrating TranSiGen into the drug discovery and mechanism research holds significant promise for advancing biomedicine.

The therapeutic effect of a novel GAPDH inhibitor in mouse model of breast cancer and efficacy monitoring by molecular imaging.

BACKGROUND: Breast cancer is a serious threat to women's health with high morbidity and mortality. The development of more effective therapies for the treatment of breast cancer is strongly warranted. Growing evidence suggests that targeting glucose metabolism may be a promising cancer treatment strategy. We previously identified a new glyceraldehyde-3-phosphate dehydrogenase (GAPDH) inhibitor, DC-5163, which shows great potential in inhibiting tumor growth. Here, we evaluated the anticancer potential of DC-5163 in breast cancer cells. METHODS: The effects of DC-5163 on breast cancer cells were investigated in vitro and in vivo. Seahorse, glucose uptake, lactate production, and cellular ATP content assays were performed to examine the impact of DC-5163 on cellular glycolysis. Cell viability, colony-forming ability, cell cycle, and apoptosis were assessed by CCK8 assay, colony formation assay, flow cytometry, and immunoblotting respectively. The anticancer activity of DC-5163 in vivo was evaluated in a mouse breast cancer xenograft model. RESULTS: DC-5163 suppressed aerobic glycolysis and reduced energy supply of breast cancer cells, thereby inhibiting breast cancer cell growth, inducing cell cycle arrest in the G0/G1 phase, and increasing apoptosis. The therapeutic efficacy was assessed using a breast cancer xenograft mouse model. DC-5163 treatment markedly suppressed tumor growth in vivo without inducing evident systemic toxicity. Micro-PET/CT scans revealed a notable reduction in tumor 18F-FDG and 18F-FLT uptake in the DC-5163 treatment group compared to the DMSO control group. CONCLUSIONS: Our results suggest that DC-5163 is a promising GAPDH inhibitor for suppressing breast cancer growth without obvious side effects. 18F-FDG and 18F-FLT PET/CT can noninvasively assess the levels of glycolysis and proliferation in tumors following treatment with DC-5163.

TG468: a text graph convolutional network for predicting clinical response to immune checkpoint inhibitor therapy.

Enhancing cancer treatment efficacy remains a significant challenge in human health. Immunotherapy has witnessed considerable success in recent years as a treatment for tumors. However, due to the heterogeneity of diseases, only a fraction of patients exhibit a positive response to immune checkpoint inhibitor (ICI) therapy. Various single-gene-based biomarkers and tumor mutational burden (TMB) have been proposed for predicting clinical responses to ICI; however, their predictive ability is limited. We propose the utilization of the Text Graph Convolutional Network (GCN) method to comprehensively assess the impact of multiple genes, aiming to improve the predictive capability for ICI response. We developed TG468, a Text GCN model framing drug response prediction as a text classification task. By combining natural language processing (NLP) and graph neural network techniques, TG468 effectively handles sparse and high-dimensional exome sequencing data. As a result, TG468 can distinguish survival time for patients who received ICI therapy and outperforms single gene biomarkers, TMB and some classical machine learning models. Additionally, TG468's prediction results facilitate the identification of immune status differences among specific patient types in the Cancer Genome Atlas dataset, providing a rationale for the model's predictions. Our approach represents a pioneering use of a GCN model to analyze exome data in patients undergoing ICI therapy and offers inspiration for future research using NLP technology to analyze exome sequencing data.

KinomeMETA: meta-learning enhanced kinome-wide polypharmacology profiling.

Kinase inhibitors are crucial in cancer treatment, but drug resistance and side effects hinder the development of effective drugs. To address these challenges, it is essential to analyze the polypharmacology of kinase inhibitor and identify compound with high selectivity profile. This study presents KinomeMETA, a framework for profiling the activity of small molecule kinase inhibitors across a panel of 661 kinases. By training a meta-learner based on a graph neural network and fine-tuning it to create kinase-specific learners, KinomeMETA outperforms benchmark multi-task models and other kinase profiling models. It provides higher accuracy for understudied kinases with limited known data and broader coverage of kinase types, including important mutant kinases. Case studies on the discovery of new scaffold inhibitors for membrane-associated tyrosine- and threonine-specific cdc2-inhibitory kinase and selective inhibitors for fibroblast growth factor receptors demonstrate the role of KinomeMETA in virtual screening and kinome-wide activity profiling. Overall, KinomeMETA has the potential to accelerate kinase drug discovery by more effectively exploring the kinase polypharmacology landscape.

Tora3D: an autoregressive torsion angle prediction model for molecular 3D conformation generation.

Three-dimensional (3D) conformations of a small molecule profoundly affect its binding to the target of interest, the resulting biological effects, and its disposition in living organisms, but it is challenging to accurately characterize the conformational ensemble experimentally. Here, we proposed an autoregressive torsion angle prediction model Tora3D for molecular 3D conformer generation. Rather than directly predicting the conformations in an end-to-end way, Tora3D predicts a set of torsion angles of rotatable bonds by an interpretable autoregressive method and reconstructs the 3D conformations from them, which keeps structural validity during reconstruction. Another advancement of our method over other conformational generation methods is the ability to use energy to guide the conformation generation. In addition, we propose a new message-passing mechanism that applies the Transformer to the graph to solve the difficulty of remote message passing. Tora3D shows superior performance to prior computational models in the trade-off between accuracy and efficiency, and ensures conformational validity, accuracy, and diversity in an interpretable way. Overall, Tora3D can be used for the quick generation of diverse molecular conformations and 3D-based molecular representation, contributing to a wide range of downstream drug design tasks.

143D, a novel selective KRASG12C inhibitor exhibits potent antitumor activity in preclinical models.

The KRASG12C mutant has emerged as an important therapeutic target in recent years. Covalent inhibitors have shown promising antitumor activity against KRASG12C-mutant cancers in the clinic. In this study, a structure-based and focused chemical library analysis was performed, which led to the identification of 143D as a novel, highly potent and selective KRASG12C inhibitor. The antitumor efficacy of 143D in vitro and in vivo was comparable with that of AMG510 and of MRTX849, two well-characterized KRASG12C inhibitors. At low nanomolar concentrations, 143D showed biochemical and cellular potency for inhibiting the effects of the KRASG12C mutation. 143D selectively inhibited cell proliferation and induced G1-phase cell cycle arrest and apoptosis by downregulating KRASG12C-dependent signal transduction. Compared with MRTX849, 143D exhibited a longer half-life and higher maximum concentration (Cmax) and area under the curve (AUC) values in mouse models, as determined by tissue distribution assays. Additionally, 143D crossed the blood‒brain barrier. Treatment with 143D led to the sustained inhibition of KRAS signaling and tumor regression in KRASG12C-mutant tumors. Moreover, 143D combined with EGFR/MEK/ERK signaling inhibitors showed enhanced antitumor activity both in vitro and in vivo. Taken together, our findings indicate that 143D may be a promising drug candidate with favorable pharmaceutical properties for the treatment of cancers harboring the KRASG12C mutation.

Structural Optimization of Fibroblast Growth Factor Receptor Inhibitors for Treating Solid Tumors.

Small-molecule fibroblast growth factor receptor (FGFR) inhibitors have emerged as a promising antitumor therapy. Herein, by further optimizing the lead compound 1 under the guidance of molecular docking, we obtained a series of novel covalent FGFR inhibitors. After careful structure-activity relationship analysis, several compounds were identified to exhibit strong FGFR inhibitory activity and relatively better physicochemical and pharmacokinetic properties compared with those of 1. Among them, 2e potently and selectively inhibited the kinase activity of FGFR1-3 wildtype and high-incidence FGFR2-N549H/K-resistant mutant kinase. Furthermore, it suppressed cellular FGFR signaling, exhibiting considerable antiproliferative activity in FGFR-aberrant cancer cell lines. In addition, the oral administration of 2e in the FGFR1-amplified H1581, FGFR2-amplified NCI-H716, and SNU-16 tumor xenograft models demonstrated potent antitumor efficacy, inducing tumor stasis or even tumor regression.

Computing the relative binding affinity of ligands based on a pairwise binding comparison network.

Structure-based lead optimization is an open challenge in drug discovery, which is still largely driven by hypotheses and depends on the experience of medicinal chemists. Here we propose a pairwise binding comparison network (PBCNet) based on a physics-informed graph attention mechanism, specifically tailored for ranking the relative binding affinity among congeneric ligands. Benchmarking on two held-out sets (provided by Schrödinger and Merck) containing over 460 ligands and 16 targets, PBCNet demonstrated substantial advantages in terms of both prediction accuracy and computational efficiency. Equipped with a fine-tuning operation, the performance of PBCNet reaches that of Schrödinger's FEP+, which is much more computationally intensive and requires substantial expert intervention. A further simulation-based experiment showed that active learning-optimized PBCNet may accelerate lead optimization campaigns by 473%. Finally, for the convenience of users, a web service for PBCNet is established to facilitate complex relative binding affinity prediction through an easy-to-operate graphical interface.

[Study on prevalence and correlation factors of bronchial asthma in Zaozhuang area, Shandong province].

OBJECTIVE: To study the prevalence of asthma and its correlated factors in Zaozhuang area in 2003, to provide a basic consideration for prevention/treatment and control policy. METHODS: 6 points were selected by stratified-clusterd-random sampling with a total of 16,725 persons expected, but only 10,610 subjects investigated. RESULTS: In this survey, 128 asthma cases were identified with a overall prevalence of 1.21%. The prevalence for children was 2.02%, and for adult was 0.90% with the former significantly higher then the latter (chi(2) = 21.39, P < 0.01). Rates for male and female were 1.08%, 1.32% with a ratio of 1:1.22. For 77.97% of children with asthma. The initiative age of asthma was before 7 years old among children while among 36.23% of the adults, it was before 15 years of age. Correlation analysis showed that upper respiratory tract infection (OR = 17.81, 95% CI: 12.25-25.89), cold air exposure (OR = 3.43, 95% CI: 2.41-4.90), stimulation through cooking and by harmful gases (OR = 2.56, 95% CI: 1.80-3.63), allergic materials (OR = 2.74, 95% CI: 1.80-4.17) were main inducing factors. 65.63% of the asthma cases having had history of allergic disease while 25.78% having had family history with the OR of allergic history and family history as 21.69 vs. 73.96. CONCLUSION: The epidemic status of bronchial asthma was serious, with an assumption that asthma cases might have reached the number of 43 thousand in Zaozhuang area.