RESUMO
Enhancers are a class of noncoding DNA, serving as crucial regulatory elements in governing gene expression by binding to transcription factors. The identification of enhancers holds paramount importance in the field of biology. However, traditional experimental methods for enhancer identification demand substantial human and material resources. Consequently, there is a growing interest in employing computational methods for enhancer prediction. In this study, we propose a two-stage framework based on deep learning, termed CapsEnhancer, for the identification of enhancers and their strengths. CapsEnhancer utilizes chaos game representation to encode DNA sequences into unique images and employs a capsule network to extract local and global features from sequence "images". Experimental results demonstrate that CapsEnhancer achieves state-of-the-art performance in both stages. In the first and second stages, the accuracy surpasses the previous best methods by 8 and 3.5%, reaching accuracies of 94.5 and 95%, respectively. Notably, this study represents the pioneering application of computer vision methods to enhancer identification tasks. Our work not only contributes novel insights to enhancer identification but also provides a fresh perspective for other biological sequence analysis tasks.
RESUMO
Application of long non-coding RNAs (lncRNAs) for modulation of breast cancer (BC) has attracted much attention. Here, we probed into the role and underlying mechanism of long intergenic non-coding RNA 01270 (LINC01270) in BC. With the help of bioinformatics tools, we identified laminin subunit alpha 2 (LAMA2) as a BC-related differentially expressed gene to discern the effect of LAMA2 in BC cells. LAMA2 was initially poorly expressed while LINC01270 was highly expressed in BC. BC cells were subsequently treated with sh-LINC01270 or/and sh-LAMA2 for exploration of their regulatory mechanism in BC, which unfolded that LINC01270 inhibition up-regulated LAMA2 and inactivated the MAPK signaling pathway to suppress malignant characteristics of BC cells. Functional assays demonstrated that LINC01270 bound to DNMT1, DNMT3a, and DNMT3b promoted the methylation of CpG islands in LAMA2 promoter and inhibited the LAMA2 expression. Moreover, our data suggested that LAMA2 suppressed MAPK signaling pathway to inhibit BC cell malignant characteristics. The in vitro results were re-produced with the help of the in vivo experimentations. In conclusion, LINC01270 silencing inhibited the methylation of LAMA2 promoter to suppress the activation of MAPK signaling pathway, which subsequently restrained the BC progression. 1, Overexpression of LAMA2 inhibits malignant features of BC cells. 2, LINC01270 promotes LAMA2 promoter methylation by recruiting DNMTs to the LAMA2 promoter region. 3, 5-aza-dc reverses the promotion of LAMA2 promoter methylation by LINC01270. 4, LAMA2 inhibits malignant features of BC cells by suppressing the activation of MAPK signaling pathway.
Assuntos
Neoplasias da Mama , RNA Longo não Codificante , Humanos , Feminino , Neoplasias da Mama/metabolismo , Epigênese Genética/genética , Metilação de DNA/genética , RNA Longo não Codificante/genética , RNA Longo não Codificante/metabolismo , Transdução de Sinais/genética , Regulação Neoplásica da Expressão Gênica/genética , Regiões Promotoras Genéticas/genética , Linhagem Celular TumoralRESUMO
Emerging evidence indicated that changes in DNA methylation early in breast cancer (BC) development might be clinically relevant for therapeutic decisions. Through analysis of whole-genome gene expression microarray and DNA methylation microarray, we explored genes with abnormal DNA methylation in BC for early detection. Firstly, human BC tissues and adjacent non-cancerous tissues were collected from nine BC patients. Gene expression microarray sequencing was conducted for identifying differentially expressed genes and DNA methylation microarray sequencing for differentially methylated genes in BC. Differentially expressed genes and methylated genes in BC were further explored using the Cancer Genome Atlas database. The correlation between DNA methylation and gene expression was illustrated by multiple comparisons. In other 60 clinical samples, methylation specific polymerase chain reaction (PCR) and reverse transcription quantitative PCR were applied for the methylation of HOXA4 and IGF1 genes in BC and adjacent non-cancerous tissues. In total, 1680 upregulated genes and 1249 downregulated genes were determined in BC. Chromosome 16 and 17 showed more differentially methylated genes, and DNA methylation level was increased in BC tissues in each gene region. Chromosome 19 showed more differentially methylated genes, and DNA methylation level was increased in BC tissues in the exoniensis 1, untranslated region-5 and transcriptional start site 200 gene regions. In other 60 clinical samples, HOXA4 and IGF1 in BC tissues presented increased DNA methylation and decreased gene expression in BC. MCF7 cells treated with RG108 showed decreased HOXA4 and IGF1 expressions. It was estimated that HOXA4 and IGF1 were identified with increased DNA methylation and decreased gene expression in BC, which may serve as biomarkers in early BC detection.
Assuntos
Neoplasias da Mama/genética , Metilação de DNA/genética , Detecção Precoce de Câncer , Genoma Humano/genética , Proteínas de Homeodomínio/genética , Fator de Crescimento Insulin-Like I/genética , Análise de Sequência com Séries de Oligonucleotídeos/métodos , Fatores de Transcrição/genética , Adulto , Idoso , Biomarcadores Tumorais/genética , Biomarcadores Tumorais/metabolismo , DNA (Citosina-5-)-Metiltransferases/antagonistas & inibidores , Metilação de DNA/efeitos dos fármacos , Bases de Dados de Ácidos Nucleicos , Feminino , Regulação Neoplásica da Expressão Gênica , Proteínas de Homeodomínio/metabolismo , Humanos , Fator de Crescimento Insulin-Like I/metabolismo , Células MCF-7 , Pessoa de Meia-Idade , Ftalimidas/farmacologia , Transdução de Sinais/efeitos dos fármacos , Fatores de Transcrição/metabolismo , Transcriptoma/genética , Triptofano/análogos & derivados , Triptofano/farmacologia , Regulação para Cima/genéticaRESUMO
BACKGROUND: It is unknown how the cell cycle plays a role in breast cancer (BC). This study aimed to establish a clinically applicable predictive model to predict the therapeutic responses and overall survival in BC patients. MATERIALS AND METHODS: Cell cycle-related genes (CCGs) were identified within the Cancer Genome Atlas cohort (n is equal to 1001) and the Gene Expression Omnibus cohort (n is equal to 3265). An analysis of univariate and multivariate Cox was then conducted to develop a nomogram based on CCGs. After validating the nomogram, risk cohort stratification was established and the predictive value was examined. Finally, immune cell infiltration and therapeutic responses were analysed. RESULTS: Based on 15 CCGs, 4 prognostic predictors were identified and entered into the nomogram. According to the curves of calibration, the estimated and observed value of the nomogram is in optimal agreement. Subsequently, stratification into two risk cohorts showed that the predictive value, immune cell infiltration and overall survival were better among patients with low risk. Immune checkpoint expression in patients with BC at higher risk was downregulated. Furthermore, the results of the study revealed that doxorubicin, paclitaxel, docetaxel, cisplatin and vinorelbine all had higher IC50 values in patients with high-risk BC. CONCLUSION: The nomogram based on CCG could assess tumour immune micro-environment regulation and therapeutic responses of immunotherapy in BC. Moreover, it may provide novel information on the control of immune micro-environment infiltration in BC and aid in the development of targeted immunotherapy.