Non-small-cell lung cancer (NSCLC) is one of the most common malignancies, and specific molecular targets are still lacking. Angiogenesis plays a central regulatory role in the growth and metastasis of malignant tumors and angiogenic factors (AFs) are involved. Although there are many studies comparing AFs and cancer, a prognostic risk model for AFs and cancer in humans has not been reported in the literature. This study aimed to identify the key AFs closely related to the process of NSCLC development, and four genes have been found, C1QTNF6, SLC2A1, PTX3, and FSTL3. Then, we constructed a novel prognostic risk model based on these four genes in non-small-cell lung cancer (NSCLC) and fully analyzed the relationship with clinical features, immune infiltration, genomes, and predictors. This model had good discrimination and calibration and will perform well in predicting the prognosis of treatment in clinical practice.
Non-small-cell lung cancer (NSCLC) is one of the most common malignancies, and specific molecular targets are still lacking. Angiogenesis plays a central regulatory role in the growth and metastasis of malignant tumor. Angiogenic factors (AFs) are involved. Although there are many studies comparing AFs and cancer, a prognostic risk model for AFs and cancer in humans has not been reported in the literature. This study aimed to identify the key AFs closely related to the process of NSCLC development, and four genes have been found, C1QTNF6, SLC2A1, PTX3, FSTL3. Then we constructed a novel prognostic risk model based on these four genes in non-small-cell lung cancer (NSCLC) and fully analyzed the relationship with clinical features, immune infiltration, genomes and predictors. This model had good discrimination and calibration and will perform well for predicting the prognosis of treatment in clinical practice.
BACKGROUND: The comprehensive data on psoriasis research are numerous and complex, making it difficult to retrieve and classify manually. The ability to quickly mine literature based on various fine topics using deep learning natural language processing technology to assess research topics and trends in the field of psoriasis disease will have a significant impact on doctors' research and patients' health education. METHOD: A neural topic model is used to identify fine topics of psoriasis literature published in the PubMed database from 2000 to 2021. Dermatologists evaluate the algorithm-modeled topics, summarize the categories into the most effective topics, and perform linear trend model analysis. The accurate classified topics are presented on an interactive web page to identify research hotspots and trends. RESULTS: At the categorical level, after review by clinicians, 158 out of 160 generated topics were found effective and categorized into 8 groups: Therapeutic methods (34.34%), pathological mechanisms (23.46%), comorbidity (20.04%), Clinical manifestations and differential diagnosis (12.77%), experimental modalities and methods (3.22%), diagnostic tools (2.99%), epidemiology (1.75%), and meetings/guidelines (1.43%). A linear regression model had good accuracy (MSE=0.252602, SSE=42.1845) and strong correlation (R-Squared=0.898009). ANOVA results showed that categories significantly impacted the model (p<=0.05), with experimental modalities and methods having the strongest relationship with year, and clinical manifestations and differential diagnosis having the weakest. An interactive web tool (https://psknlr.github.io) facilitates quick retrieval of titles, journals, and abstracts in different categories, as well as browsing literature information under specific topics and accessing corresponding article pages for professional knowledge on psoriasis. CONCLUSIONS: The neural topic model and interactive web tool can effectively identify the research hotspots and trends in psoriasis literature, assisting clinicians and patients in retrieving and comparing pertinent topics and research accomplishments of various years. Keywords: psoriasis; topic model; PubMed; deep learning; pre-trained language model
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.