RESUMO
This study has used machine learning algorithms to develop a predictive model for differentiating between dermoscopic images of basal cell carcinoma (BCC) and actinic keratosis (AK). We compiled a total of 904 dermoscopic images from two sources - the public dataset (HAM10000) and our proprietary dataset from the First Affiliated Hospital of Dalian Medical University (DAYISET 1) - and subsequently categorised these images into four distinct cohorts. The study developed a deep learning model for quantitative analysis of image features and integrated 15 machine learning algorithms, generating 207 algorithmic combinations through random combinations and cross-validation. The final predictive model, formed by integrating XGBoost with Lasso regression, exhibited effective performance in the differential diagnosis of BCC and AK. The model demonstrated high sensitivity in the training set and maintained stable performance in three validation sets. The area under the curve (AUC) value reached 1.000 in the training set and an average of 0.695 in the validation sets. The study concludes that the constructed discriminative diagnostic model based on machine learning algorithms has excellent predictive capabilities that could enhance clinical decision-making efficiency, reduce unnecessary biopsies, and provide valuable guidance for further treatment.
RESUMO
Brusatol (Bru), a main extract from traditional Chinese medicine Brucea javanica, has been reported to exist antitumor effect in many tumors including melanoma. However, the underlying mechanism in its anti-melanoma effect still need further exploration. Here, we reported that the protein expression of KLF4 in melanoma cells were significantly downregulated in response to brusatol treatment. Overexpression of KLF4 suppressed brusatol-induced melanoma cell apoptosis; while knockdown of KLF4 enhanced antitumor effects of brusatol on melanoma cells not only in vitro but also in vivo. Further studies on the mechanism revealed that KLF4 bound to the promoter of NCK2 directly and facilitated NCK2 transcription, which suppressed the antitumor effect of brusatol on melanoma. Furthermore, our findings showed that miR-150-3p was dramatically upregulated under brusatol treatment which resulted in the downregulation of KLF4. Our results suggested that the miR-150-3p/KLF4/NCK2 axis might play an important role in the antitumour effects of brusatol in melanoma.
Assuntos
Melanoma , MicroRNAs , Quassinas , Humanos , Melanoma/tratamento farmacológico , Melanoma/genética , Melanoma/metabolismo , Quassinas/farmacologia , Apoptose , MicroRNAs/genética , MicroRNAs/farmacologia , Proteínas Oncogênicas/metabolismo , Proteínas Adaptadoras de Transdução de Sinal/genética , Proteínas Adaptadoras de Transdução de Sinal/metabolismoRESUMO
Melanoma is a malignant tumor of melanocytes and is often considered immunogenic cancer. Toll-like receptor-related genes are expressed differently in most types of cancer, depending on the immune microenvironment inside cancer, and the key function of Toll-like receptors (TLRs) for melanoma has not been fully elucidated. Based on multi-omics data from TCGA and GEO databases, we first performed pan-cancer analysis on TLR, including CNV, SNV, and mRNA changes in TLR-related genes in multiple human cancers, as well as patient prognosis characterization. Then, we divided melanoma patients into three subgroups (clusters 1, 2, and 3) according to the expression of the TLR pathway, and explored the correlation between TLR pathway and melanoma prognosis, immune infiltration, metabolic reprogramming, and oncogene expression characteristics. Finally, through univariate Cox regression analysis and LASSO algorithm, we selected six TLR-related genes to construct a survival prognostic model, divided melanoma patients into the training set, internal validation set 1, internal validation set 2, and external validation set for multiple validations, and discussed the correlation between model genes and clinical features of melanoma patients. In conclusion, we constructed a prognostic survival model based on TLR-related genes that precisely and independently demonstrated the potential to assess the prognosis and immune traits of melanoma patients, which is critical for patients' survival.
Assuntos
Melanoma , Humanos , Melanoma/genética , Oncogenes , Melanócitos , Algoritmos , Transdução de Sinais/genética , Microambiente Tumoral/genéticaRESUMO
Brusatol (Bru), a Chinese medicine Brucea javanica extract, has a variety of antitumour effects. However, its role and underlying mechanism in melanoma have not been fully elucidated. In this study, we found that brusatol inhibited melanoma cell proliferation and migration and promoted cell apoptosis in vitro, in addition to suppressing melanoma cell tumorigenesis in vivo. Further studies on the mechanism revealed that brusatol significantly downregulated the expression of stearoyl-CoA desaturase 1 (SCD1). Increased SCD1 expression could impair the antitumour effects of brusatol on melanoma cells. Subsequently, we found that HOXB9, an important transcription factor, was directly bound to the promoter of SCD1, facilitating its transcription. Overexpression of HOXB9 inhibited brusatol-induced SCD1 reduction and promoted cell survival. Furthermore, our results revealed that miR-122-5p was significantly increased in response to brusatol treatment and led to a decrease in HOXB9 in melanoma. Collectively, our data suggested that the miR-122-5p/HOXB9/SCD1 axis might play an important role in the antitumour effects of brusatol and that brusatol might have potential clinical implications in melanoma therapy.
Assuntos
Melanoma , MicroRNAs , Quassinas , Humanos , Melanoma/patologia , Regulação da Expressão Gênica , MicroRNAs/genética , Linhagem Celular Tumoral , Proteínas de Homeodomínio/genética , Estearoil-CoA Dessaturase/genéticaRESUMO
Skin cutaneous melanoma (SKCM) is a common cancer of which mortality is increasing continuously. Our study conducted a series of analyses on the clinical significance of Serine/threonine kinase 17B (STK17B) in SKCM to provide a new biomarker for diagnosis and treatment. The RNA-sequence data were obtained from The Cancer Genome Atlas and Genotype-Tissue Expression databases. The data of 468 SKCM patients were divided into STK17B high- and low-expression groups and analyzed by Bioconductor package to identify the differential expressed genes. The R package of "clusterProfiler" was used for Gene Ontology, Kyoto Encyclopedia of Genes and Genomes, and Gene-Set Enrichment Analysis analyses. A protein-protein interaction network and immune infiltration landscape were respectively constructed via STRING database and ssGSEA. STK17B had lower expression in SKCM than normal tissues. Besides, STK17B expression was significantly related to some clinicopathological characteristics in SKCM patients including T stage, Breslow depth, radiation therapy, melanoma Clark level, and pathologic stage. The Kaplan-Meier curve analyses revealed that the low expression of STK17B was correlated with poor overall survival and disease-specific survival. We constructed nomograms to predict the 1-, 3-, and 5-year survival of SKCM patients. The function enrichment analyses showed STK17B-related differential expressed genes were enriched in cellular differentiation and immune-related progress. STK17B expression level were positively correlated with infiltrating level of immune cells. In this study, we found that STK17B, which played an important role in immune infiltration, could be a new biomarker for diagnosis and prognosis in SKCM patients.
Assuntos
Proteínas Reguladoras de Apoptose/genética , Biomarcadores Tumorais/genética , Melanoma/patologia , Nomogramas , Proteínas Serina-Treonina Quinases/genética , Neoplasias Cutâneas/patologia , Microambiente Tumoral , Idoso , Estudos de Casos e Controles , Bases de Dados Genéticas , Feminino , Seguimentos , Humanos , Masculino , Melanoma/genética , Melanoma/imunologia , Melanoma/radioterapia , Pessoa de Meia-Idade , Prognóstico , Neoplasias Cutâneas/genética , Neoplasias Cutâneas/imunologia , Neoplasias Cutâneas/radioterapia , Taxa de Sobrevida , Melanoma Maligno CutâneoRESUMO
Long non-coding RNAs (lncRNAs) are a kind of translational-repressor RNAs composed of more than 200 nucleotides and formerly considered as "transcriptional noise". Recently studies have shown that lncRNAs could bind to multiple biomolecules such as DNA, transcription factors, RNA, chromatin complexes and proteins, and regulate target gene expression at multi-levels, thus playing an essential role in human tumors. DLX6-AS1, a recently discovered oncogenic lncRNA, is highly expressed in various human tumors, including lung cancer, liver cancer and pancreatic cancer. This paper mainly reviewed the regulatory mechanism of DLX6-AS1 as a competitive endogenous RNA (ceRNA) in tumor cell proliferation, cell apoptosis, angiogenesis, epithelial-mesenchymal transformation, chemotherapy resistance and metabolic changes. Furthermore, the translational value of DLX6-AS1 in cancer was also elucidated, which suggested its potential as a diagnostic or prognostic biomarker in cancer. In summary, this present article not only makes an in-depth analysis of the expression changes and carcinogenic mechanism of DLX6-AS1 in various human cancers, but also provides a new breakthrough for the diagnosis and treatment of cancers.
Assuntos
Biomarcadores Tumorais , Regulação Neoplásica da Expressão Gênica/genética , Proteínas de Homeodomínio/genética , Neoplasias/genética , Neoplasias/patologia , Apoptose/genética , Proliferação de Células/genética , Resistencia a Medicamentos Antineoplásicos/genética , Transição Epitelial-Mesenquimal/genética , Expressão Gênica , Proteínas de Homeodomínio/metabolismo , Humanos , Terapia de Alvo Molecular , Neoplasias/diagnóstico , Neoplasias/terapia , Neovascularização Patológica/patologia , Prognóstico , Biossíntese de ProteínasRESUMO
Uveal melanoma (UVM), the most common primary intraocular malignancy, has a high mortality because of a high propensity to metastasize. Our study analyzed prognostic value and immune-related characteristics of CARD11 in UVM, hoping to provide a potential management and research direction. The RNA-sequence data of 80 UVM patients were downloaded from The Cancer Genome Atlas database and divided them into high- and low-expression groups. We analyzed the differentially expressed genes, enrichment analyses and the infiltration of immune cells using the R package and Gene-Set Enrichment Analysis. A clinical prediction nomogram and protein-protein interaction network were constructed and the first 8 genes were considered as the hub-genes. Finally, we constructed a competing endogenous RNA (ceRNA) network by Cytoscape and analyzed the statistical data via the R software. Here we found that CARD11 expression had notable correlation with UVM clinicopathological features, which was also an independent predictor for overall survival (OS). Intriguingly, CARD11 had a positively correlation to autophagy, cellular senescence and apoptosis. Infiltration of monocytes was significantly higher in low CARD11 expression group, and infiltration of T cells regulatory was lower in the same group. Functional enrichment analyses revealed that CARD11 was positively related to T cell activation pathways and cell adhesion molecules. The expressions of hub-genes were all increased in the high CARD11 expression group and the ceRNA network showed the interaction among mRNA, miRNA and lncRNA. These findings show that high CARD11 expression in UVM is associated with poor OS, indicating that CARD11 may serve as a potential biomarker for the diagnosis and prognosis of the UVM.