A deep learning based multimodal fusion model for skin lesion diagnosis using smartphone collected clinical images and metadata.

Introduction: Skin cancer is one of the most common types of cancer. An accessible tool to the public can help screening for malign lesion. We aimed to develop a deep learning model to classify skin lesion using clinical images and meta information collected from smartphones. Methods: A deep neural network was developed with two encoders for extracting information from image data and metadata. A multimodal fusion module with intra-modality self-attention and inter-modality cross-attention was proposed to effectively combine image features and meta features. The model was trained on tested on a public dataset and compared with other state-of-the-art methods using five-fold cross-validation. Results: Including metadata is shown to significantly improve a model's performance. Our model outperformed other metadata fusion methods in terms of accuracy, balanced accuracy and area under the receiver-operating characteristic curve, with an averaged value of 0.768±0.022, 0.775±0.022 and 0.947±0.007. Conclusion: A deep learning model using smartphone collected images and metadata for skin lesion diagnosis was successfully developed. The proposed model showed promising performance and could be a potential tool for skin cancer screening.

TIMM13 as a prognostic biomarker and associated with immune infiltration in skin cutaneous melanoma (SKCM).

Objective: Providing protection against aggregation and guiding hydrophobic precursors through the mitochondria's intermembrane space, this protein functions as a chaperone-like protein. SLC25A12 is imported by TIMM8 as a result of its interaction with TIMM13. In spite of this, it is still unknown how TIMM13 interacts with skin cutaneous melanoma (SKCM) and tumor-infiltrating lymphocytes (TILs). Methods: Aberrant expression of TIMM13 in SKCM and its clinical outcome was evaluated with the help of multiple databases, including the Xiantao tool (https://www.xiantao.love/), HPA, and UALCAN. TISIDB and Tumor Immune Estimation Resources (TIMER) databases were applied to explore the association between TIMM13 and tumor infiltration immune cells. OS nomogram was constructed, and model performance was examined. Finally, TIMM13 protein expression was validated by immunohistochemistry (IHC). Results: TIMM13 expression was higher in SKCM samples than in peritumor samples. TIMM13 was strongly associated with sample type, subgroup, cancer stage, lymph node stage, and worse survival. Further, upregulation of TIMM13 was significantly associated with immunoregulators, and chemokines, as well as T cells, B cells, monocytes, neutrophils, macrophages, and T-cell regulators. An analysis of bioinformatic data uncovered that TIMM13 expression was strongly associated with PD1 (T-cell exhaustion marker). The nomogram showed good predictive performance based on calibration plot. TIMM13 was highly expressed in melanoma tissue samples than in normal samples. Conclusion: In brief, TIMM13 may be a prognostic biomarker for SKCM. It might modulate the tumor immune microenvironment and lead to a poorer prognosis. In addition, it is necessary to study the targeted therapy of TIMM13.

Identification and characterization of four immune-related signatures in keloid.

A keloid is a fibroproliferative disorder of unknown etiopathogenesis that requires ill-defined treatment. Existing evidence indicates that the immune system plays an important role in the occurrence and development of keloid. However, there is still a lack of research on the immune-related signatures of keloid. Here we identified immune-related signatures in keloid and explored their pathological mechanisms. Transcriptomic datasets (GSE7890, GSE92566, and GSE44270) of keloid and normal skin tissues were obtained from the Gene Expression Omnibus database. The overlap of differentially expressed genes and immune-related genes was considered as differentially expressed immune-related genes (DEIGs). Functional analysis, expression, and distribution were applied to explore the function and characteristics of DEIGs, and the expression of these DEIGs in keloid and normal skin tissues was verified by immunohistochemistry. Finally, we conducted interactive network analysis and immune infiltration analysis to determine the therapeutic potential and immune correlation. We identified four DEIGs (LGR5, PTN, JAG1, and DKK1). In these datasets, only GSE7890 met the screening criteria. In the GSE7890 dataset, DKK1 and PTN were downregulated in keloid, whereas JAG1 and LGR5 were upregulated in keloid. In addition, we obtained the same conclusion through immunohistochemistry. Functional analysis indicated that these four DEIGs were mainly involved in stem cell, cell cycle, UV response, and therapy resistance. Through interactive network analysis, we found that these DEIGs were associated with drugs currently used to treat keloid, such as hydrocortisone, androstanolone, irinotecan, oxaliplatin, BHQ-880, and lecoleucovorin. Finally, many immune cells, including CD8+ T cells, resting memory CD4+ T cells, and M1 macrophages, were obtained by immune infiltration analysis. In conclusion, we identified four immune signaling molecules associated with keloid (LGR5, PTN, JAG1, and DKK1). These immune-related signaling molecules may be important modules in the pathogenesis of keloid. Additionally, we developed novel therapeutic targets for the treatment of this challenging disease.

New Prognostic Biomarkers and Drug Targets for Skin Cutaneous Melanoma via Comprehensive Bioinformatic Analysis and Validation.

Skin cutaneous melanoma (SKCM) is the most aggressive and fatal type of skin cancer. Its highly heterogeneous features make personalized treatments difficult, so there is an urgent need to identify markers for early diagnosis and therapy. Detailed profiles are useful for assessing malignancy potential and treatment in various cancers. In this study, we constructed a co-expression module using expression data for cutaneous melanoma. A weighted gene co-expression network analysis was used to discover a co-expression gene module for the pathogenesis of this disease, followed by a comprehensive bioinformatics analysis of selected hub genes. A connectivity map (CMap) was used to predict drugs for the treatment of SKCM based on hub genes, and immunohistochemical (IHC) staining was performed to validate the protein levels. After discovering a co-expression gene module for the pathogenesis of this disease, we combined GWAS validation and DEG analysis to identify 10 hub genes in the most relevant module. Survival curves indicated that eight hub genes were significantly and negatively associated with overall survival. A total of eight hub genes were positively correlated with SKCM tumor purity, and 10 hub genes were negatively correlated with the infiltration level of CD4+ T cells and B cells. Methylation levels of seven hub genes in stage 2 SKCM were significantly lower than those in stage 3. We also analyzed the isomer expression levels of 10 hub genes to explore the therapeutic target value of 10 hub genes in terms of alternative splicing (AS). All 10 hub genes had mutations in skin tissue. Furthermore, CMap analysis identified cefamandole, ursolic acid, podophyllotoxin, and Gly-His-Lys as four targeted therapy drugs that may be effective treatments for SKCM. Finally, IHC staining results showed that all 10 molecules were highly expressed in melanoma specimens compared to normal samples. These findings provide new insights into SKCM pathogenesis based on multi-omics profiles of key prognostic biomarkers and drug targets. GPR143 and SLC45A2 may serve as drug targets for immunotherapy and prognostic biomarkers for SKCM. This study identified four drugs with significant potential in treating SKCM patients.

Identification of a novel tumour microenvironment-based prognostic biomarker in skin cutaneous melanoma.

Skin cutaneous melanoma (SKCM) is one of the most destructive skin malignancies and has attracted worldwide attention. However, there is a lack of prognostic biomarkers, especially tumour microenvironment (TME)-based prognostic biomarkers. Therefore, there is an urgent need to investigate the TME in SKCM, as well as to identify efficient biomarkers for the diagnosis and treatment of SKCM patients. A comprehensive analysis was performed using SKCM samples from The Cancer Genome Atlas and normal samples from Genotype-Tissue Expression. TME scores were calculated using the ESTIMATE algorithm, and differential TME scores and differentially expressed prognostic genes were successively identified. We further identified more reliable prognostic genes via least absolute shrinkage and selection operator regression analysis and constructed a prognostic prediction model to predict overall survival. Receiver operating characteristic analysis was used to evaluate the diagnostic efficacy, and Cox regression analysis was applied to explore the relationship with clinicopathological characteristics. Finally, we identified a novel prognostic biomarker and conducted a functional enrichment analysis. After considering ESTIMATEScore and tumour purity as differential TME scores, we identified 34 differentially expressed prognostic genes. Using least absolute shrinkage and selection operator regression, we identified seven potential prognostic biomarkers (SLC13A5, RBM24, IGHV3OR16-15, PRSS35, SLC7A10, IGHV1-69D and IGHV2-26). Combined with receiver operating characteristic and regression analyses, we determined PRSS35 as a novel TME-based prognostic biomarker in SKCM, and functional analysis enriched immune-related cells, functions and signalling pathways. Our study indicated that PRSS35 could act as a potential prognostic biomarker in SKCM by investigating the TME, so as to provide new ideas and insights for the clinical diagnosis and treatment of SKCM.

The Identification of the Metabolism Subtypes of Skin Cutaneous Melanoma Associated With the Tumor Microenvironment and the Immunotherapy.

Skin cutaneous melanoma (SKCM) is a highly aggressive and resistant cancer with immense metabolic heterogeneity. Here, we performed a comprehensive examination of the diverse metabolic signatures of SKCM based on non-negative matrix factorization (NMF) categorization, clustering SKCM into three distinct metabolic subtypes (C1, C2, and C3). Next, we evaluated the metadata sets of the metabolic signatures, prognostic values, transcriptomic features, tumor microenvironment signatures, immune infiltration, clinical features, drug sensitivity, and immunotherapy response of the subtypes and compared them with those of prior publications for classification. Subtype C1 was associated with high metabolic activity, low immune scores, and poor prognosis. Subtype C2 displayed low metabolic activity, high immune infiltration, high stromal score, and high expression of immune checkpoints, demonstrating the drug sensitivity to PD-1 inhibitors. The C3 subtype manifested moderate metabolic activity, high enrichment in carcinogenesis-relevant pathways, high levels of CpG island methylator phenotype (CIMP), and poor prognosis. Eventually, a 90-gene classifier was produced to implement the SKCM taxonomy and execute a consistency test in different cohorts to validate its reliability. Preliminary validation was performed to ascertain the role of SLC7A4 in SKCM. These results indicated that the 90-gene signature can be replicated to stably identify the metabolic classification of SKCM. In this study, a novel SKCM classification approach based on metabolic gene expression profiles was established to further understand the metabolic diversity of SKCM and provide guidance on precisely targeted therapy to patients with the disease.

UTRN inhibits melanoma growth by suppressing p38 and JNK/c-Jun signaling pathways.

BACKGROUND: Utrophin (UTRN), as a tumor suppressor gene, is involved in various cancer progression. The function of UTRN in the melanoma process and the related molecular mechanisms are still unclear. Herein, we studied the function of UTRN in melanoma growth and the relevant molecular mechanisms. METHODS: Using the GEO database and UCSC Xena project, we compared the expression of UTRN in non-cancerous and melanoma tissues. Immunohistochemistry (IHC) staining, qRT-PCR and Western Blot (WB) were performed to evaluate UTRN expression in clinical samples. A total of 447 cases with UTRN expression data, patient characteristics and survival data were extracted from TCGA database and analyzed. After stable transduction and single cell cloning, the proliferation ability of A375 human melanoma cells was analyzed by Cell Counting Kit­8 (CCK) and 5­ethynyl­2'­deoxyuridine (EdU) incorporation assays. GSEA was performed to predict the mechanism by which UTRN regulated melanoma growth. Then WB analysis was used to assess the protein expression levels of pathway signaling in overexpression (EXP) melanoma cells. Epac activator 8-pCPT-2'-O-Me-cAMP was then used to evaluate the proliferation ability by activation of p38 and JNK/c-Jun signaling pathways. RESULTS: Data from GEO and UCSC Xena project indicated that UTRN expression was decreased in melanoma. Experiment on clinical samples further confirmed our finding. TCGA results showed that a reduced expression of UTRN in 447 melanoma samples was associated with advanced clinical characteristics (T stage, Clark level, ulceration), shorter survival time and poorer prognosis. In addition, up-regulated UTRN expression inhibited melanoma cell proliferation when compared to control group. MAPK signaling pathway was presented in both KEGG and BioCarta databases by using GSEA tool. WB results confirmed the down-regulated expression of p38, JNK1 and c-Jun in EXP group when compared to control group. Epac activator 8-pCPT-2'-O-Me-cAMP treatment could partially rescue proliferation of tumor cells. CONCLUSION: We have demonstrated that reduced UTRN predicted poorer prognosis and UTRN inhibited melanoma growth via p38 and JNK1/c-Jun pathways. Therefore, UTRN could serve as a tumor suppressor and novel prognostic biomarker for melanoma patients.