A support vector machine and a random forest classifier indicates a 15-miRNA set related to osteosarcoma recurrence

GeneAI 3.0: powerful, novel, generalized hybrid and ensemble deep learning frameworks for miRNA species classification of stationary patterns from nucleotides

Due to the intricate relationship between the small non-coding ribonucleic acid (miRNA) sequences, the classification of miRNA species, namely Human, Gorilla, Rat, and Mouse is challenging. Previous methods are not robust and accurate. In this study, we present AtheroPoint's GeneAI 3.0, a powerful, novel, and generalized method for extracting features from the fixed patterns of purines and pyrimidines in each miRNA sequence in ensemble paradigms in machine learning (EML) and convolutional neural network (CNN)-based deep learning (EDL) frameworks. GeneAI 3.0 utilized five conventional (Entropy, Dissimilarity, Energy, Homogeneity, and Contrast), and three contemporary (Shannon entropy, Hurst exponent, Fractal dimension) features, to generate a composite feature set from given miRNA sequences which were then passed into our ML and DL classification framework. A set of 11 new classifiers was designed consisting of 5 EML and 6 EDL for binary/multiclass classification. It was benchmarked against 9 solo ML (SML), 6 solo DL (SDL), 12 hybrid DL (HDL) models, resulting in a total of 11 + 27 = 38 models were designed. Four hypotheses were formulated and validated using explainable AI (XAI) as well as reliability/statistical tests. The order of the mean performance using accuracy (ACC)/area-under-the-curve (AUC) of the 24 DL classifiers was: EDL > HDL > SDL. The mean performance of EDL models with CNN layers was superior to that without CNN layers by 0.73%/0.92%. Mean performance of EML models was superior to SML models with improvements of ACC/AUC by 6.24%/6.46%. EDL models performed significantly better than EML models, with a mean increase in ACC/AUC of 7.09%/6.96%. The GeneAI 3.0 tool produced expected XAI feature plots, and the statistical tests showed significant p-values. Ensemble models with composite features are highly effective and generalized models for effectively classifying miRNA sequences.

Nonnegative matrix factorization analysis and multiple machine learning methods identified IL17C and ACOXL as novel diagnostic biomarkers for atherosclerosis

BACKGROUND

Atherosclerosis is the common pathological basis for many cardiovascular and cerebrovascular diseases. The purpose of this study is to identify the diagnostic biomarkers related to atherosclerosis through machine learning algorithm.

METHODS

Clinicopathological parameters and transcriptomics data were obtained from 4 datasets (GSE21545, GSE20129, GSE43292, GSE100927). A nonnegative matrix factorization algorithm was used to classify arteriosclerosis patients in GSE21545 dataset. Then, we identified prognosis-related differentially expressed genes (DEGs) between the subtypes. Multiple machine learning methods to detect pivotal markers. Discrimination, calibration and clinical usefulness of the predicting model were assessed using area under curve, calibration plot and decision curve analysis respectively. The expression level of the feature genes was validated in GSE20129, GSE43292, GSE100927.

RESULTS

2 molecular subtypes of atherosclerosis was identified, and 223 prognosis-related DEGs between the 2 subtypes were identified. These genes are not only related to epithelial cell proliferation, mitochondrial dysfunction, but also to immune related pathways. Least absolute shrinkage and selection operator, random forest, support vector machine- recursive feature elimination show that IL17C and ACOXL were identified as diagnostic markers of atherosclerosis. The prediction model displayed good discrimination and good calibration. Decision curve analysis showed that this model was clinically useful. Moreover, IL17C and ACOXL were verified in other 3 GEO datasets, and also have good predictive performance.

CONCLUSION

IL17C and ACOXL were diagnostic genes of atherosclerosis and associated with higher incidence of ischemic events.

Identification of Markers for Diagnosis and Treatment of Diabetic Kidney Disease Based on the Ferroptosis and Immune

BACKGROUND

In advanced diabetic kidney disease (DKD), iron metabolism and immune dysregulation are abnormal, but the correlation is not clear. Therefore, we aim to explore the potential mechanism of ferroptosis-related genes in DKD and their relationship with immune inflammatory response and to identify new diagnostic biomarkers to help treat and diagnose DKD.

METHODS

Download data from gene expression omnibus (GEO) database and FerrDb database, and construct random forest tree (RF) and support vector machine (SVM) model to screen hub ferroptosis genes (DE-FRGs). We used consistent unsupervised consensus clustering to cluster DKD samples, and enrichment analysis was performed by Gene Set Variation Analysis (GSVA), Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) and then assessed immune cell infiltration abundance using the single-sample gene set enrichment analysis (ssGSEA) and CIBERSORT algorithms. Ferroptosis scoring system was established based on the Boruta algorithm, and then, core compounds were screened, and binding sites were predicted by Coremine Medical database.

RESULTS

We finally established a 7-gene signature (DUSP1, PRDX6, PEBP1, ZFP36, GABARAPL1, TSC22D3, and RGS4) that exhibited good stability across different datasets. Consistent clustering analysis divided the DKD samples into two ferroptosis modification patterns. Meanwhile, autophagy and peroxisome pathways and immune-related pathways can participate in the regulation of ferroptosis modification patterns. The abundance of immune cell infiltration differs significantly across patterns. Further, molecular docking results showed that the core compound could bind to the protein encoded by the core gene.

CONCLUSIONS

Our findings suggest that ferroptosis modification plays a crucial role in the diversity and complexity of the DKD immune microenvironment, and the ferroptosis score system can be used to effectively verify the relationship between ferroptosis and immune cell infiltration in DKD patients. Kaempferol and quercetin may be potential drugs to improve the immune and inflammatory mechanisms of DKD by affecting ferroptosis.

Identification of key snoRNAs serves as biomarkers for hepatocellular carcinoma by bioinformatics methods

BACKGROUND

Hepatocellular carcinoma (HCC) is a common malignancy with high mortality and poor prognosis due to a lack of predictive markers. However, research on small nuclear RNAs (snoRNAs) in HCC were very little. This study aimed to identify a potential diagnostic and prognostic snoRNA signature for HCC.

METHODS

HCC datasets from the cancer genome atlas (TCGA) and international cancer genome consortium (ICGC) cohorts were used. Differentially expressed snoRNA (DEs) were identified using the limma package. Based on the DEs, diagnostic and prognostic models were established by the least absolute shrinkage and selection operator (LASSO) regression and COX analysis, and Kaplan-Meier (K-M) survival analysis and receiver operating characteristic (ROC) curve analysis were conducted to evaluate the efficiency of signatures. Gene set enrichment analysis (GSEA) and gene set variation analysis (GSVA) were used to analyze the risk score and further explore the potential correlation between the risk groups and tumor immune status in TCGA. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to determine the functions of key snoRNAs.

RESULTS

We constructed a 6-snoRNAs signature which could classify patients into high- or low-risk groups and found that patients in the high-risk group had a worse prognosis than those in the low-risk group and were significantly involved in p53 processes. Tumor immune status analysis revealed that CTLA4 and PDCD1 (PD1) were highly expressed in the high-risk group, which responded to PD1 inhibitor therapy. Additionally, a 25-snoRNAs diagnostic signature was constructed with an area under the curve (AUC) of 0.933 for distinguishing HCCs from normal controls. Finally, 3 key snoRNAs (SNORA11, SNORD124, and SNORD46) were identified with both diagnostic and prognostic efficacy, some of which were closely related to the spliceosome and Notch signaling pathways.

CONCLUSIONS

Our study identified 6 snoRNAs that may serve as novel prognostic models and 3 key snoRNAs with both diagnostic and prognostic efficacy for HCC.

Construction of genetic classification model for coronary atherosclerosis heart disease using three machine learning methods

Background

Although the diagnostic method for coronary atherosclerosis heart disease (CAD) is constantly innovated, CAD in the early stage is still missed diagnosis for the absence of any symptoms. The gene expression levels varied during disease development; therefore, a classifier based on gene expression might contribute to CAD diagnosis. This study aimed to construct genetic classification models for CAD using gene expression data, which may provide new insight into the understanding of its pathogenesis.

Methods

All statistical analysis was completed by R 3.4.4 software. Three raw gene expression datasets (GSE12288, GSE7638 and GSE66360) related to CAD were downloaded from the Gene Expression Omnibus database and included for analysis. Limma package was performed to identify differentially expressed genes (DEGs) between CAD samples and healthy controls. The WGCNA package was conducted to recognize CAD-related gene modules and hub genes, followed by recursive feature elimination analysis to select the optimal features genes (OFGs). The genetic classification models were established using support vector machine (SVM), random forest (RF) and logistic regression (LR), respectively. Further validation and receiver operating characteristic (ROC) curve analysis were conducted to evaluate the classification performance.

Results

In total, 374 DEGs, eight gene modules, 33 hub genes and 12 OFGs (HTR4, KISS1, CA12, CAMK2B, KLK2, DDC, CNGB1, DERL1, BCL6, LILRA2, HCK, MTF2) were identified. ROC curve analysis showed that the accuracy of SVM, RF and LR were 75.58%, 63.57% and 63.95% in validation; with area under the curve of 0.813 (95% confidence interval, 95% CI 0.761–0.866, P < 0.0001), 0.727 (95% CI 0.665–0.788, P < 0.0001) and 0.783 (95% CI 0.725–0.841, P < 0.0001), respectively.

Conclusions

In conclusion, this study found 12 gene signatures involved in the pathogenic mechanism of CAD. Among the CAD classifiers constructed by three machine learning methods, the SVM model has the best performance.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12872-022-02481-4.

Application of artificial intelligence in the dental field: A literature review

PURPOSE

The purpose of this study was to comprehensively review the literature regarding the application of artificial intelligence (AI) in the dental field,focusing on the evaluation criteria and architecture types.

STUDY SELECTION

Electronic databases (PubMed, Cochrane Library, Scopus) were searched. Full-text articles describing the clinical application of AI for the detection, diagnosis, and treatment of lesions and the AI method/architecture were included.

RESULTS

The primary search presented 422 studies from 1996 to 2019, and 58 studies were finally selected. Regarding the year of publication, the oldest study, which was reported in 1996, focused on "oral and maxillofacial surgery." Machine-learning architectures were employed in the selected studies, while approximately half of them (29/58) employed neural networks. Regarding the evaluation criteria, eight studies compared the results obtained by AI with the diagnoses formulated by dentists, while several studies compared two or more architectures in terms of performance. The following parameters were employed for evaluating the AI performance: accuracy, sensitivity, specificity, mean absolute error, root mean squared error, and area under the receiver operating characteristic curve.

CONCLUSIONS

Application of AI in the dental field has progressed; however, the criteria for evaluating the efficacy of AI have not been clarified. It is necessary to obtain better quality data for machine learning to achieve the effective diagnosis of lesions and suitable treatment planning.

Integrated analysis of circRNA-miRNA-mRNA network reveals potential prognostic biomarkers for radiotherapies with X-rays and carbon ions in non-small cell lung cancer

Background

This work was aimed at exploring the regulatory network of non-coding RNA (ncRNA) especially circular RNA (circRNA) and microRNA (miRNA), in the sensitivity of non-small cell lung cancer (NSCLC) cells to low linear energy transfer (LET) X-ray and high-LET carbon ion irradiations.

Methods

The radioresistant NSCLC cell line A549-R11 was obtained from its parental cell line A549 through irradiation with X-rays of 2.0 Gy per fraction for 30 times. The sensitivities of A549, A549-R11 and H1299 cells exposed to X-rays and carbon ions were verified using the colony formation assay. A comprehensive circRNA-miRNA-mRNA network was constructed through the sequencing data in parental A549, acquired radioresistant A549-R11 and intrinsic radioresistant H1299 cells, and the network was further optimized according to the prognostic results from the TCGA and GEO databases.

Results

Based on high-throughput sequencing of circRNAs, we found that 40 circRNAs were up-regulated while 184 circRNAs were down-regulated in the intersection of the sets of A549-R11 and H1299 cells. Subsequently, a circRNA- miRNA-mRNA network, including 14 interactive pairs and 8 circRNAs, 4 overall survival-associated miRNAs, and 4 mRNAs, was constructed through the high-throughput data screening and bioinformatics methods.

Conclusions

Our results provide a complete understanding to the regulatory mechanism of the sensitivities to low-LET X-ray and high-LET carbon ion irradiations, and might be helpful to screen potential biomarkers for predicting the Carbon-ion radiotherapy (CIRT) and X-ray radiotherapy responses in NSCLC.

A novel risk score model based on eight genes and a nomogram for predicting overall survival of patients with osteosarcoma

BACKGROUND

This study aims to identify a predictive model to predict survival outcomes of osteosarcoma (OS) patients.

METHODS

A RNA sequencing dataset (the training set) and a microarray dataset (the validation set) were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) database, respectively. Differentially expressed genes (DEGs) between metastatic and non-metastatic OS samples were identified in training set. Prognosis-related DEGs were screened and optimized by support vector machine (SVM) recursive feature elimination. A SVM classifier was built to classify metastatic and non-metastatic OS samples. Independent prognosic genes were extracted by multivariate regression analysis to build a risk score model followed by performance evaluation in two datasets by Kaplan-Meier (KM) analysis. Independent clinical prognostic indicators were identified followed by nomogram analysis. Finally, functional analyses of survival-related genes were conducted.

RESULT

Totally, 345 DEGs and 45 prognosis-related genes were screened. A SVM classifier could distinguish metastatic and non-metastatic OS samples. An eight-gene signature was an independent prognostic marker and used for constructing a risk score model. The risk score model could separate OS samples into high and low risk groups in two datasets (training set: log-rank p < 0.01, C-index = 0.805; validation set: log-rank p < 0.01, C-index = 0.797). Tumor metastasis and RS model status were independent prognostic factors and nomogram model exhibited accurate survival prediction for OS. Additionally, functional analyses of survival-related genes indicated they were closely associated with immune responses and cytokine-cytokine receptor interaction pathway.

CONCLUSION

An eight-gene predictive model and nomogram were developed to predict OS prognosis.

Expression scoring of a small-nucleolar-RNA signature identified by machine learning serves as a prognostic predictor for head and neck cancer

Head and neck squamous cell carcinoma (HNSCC) is a common malignancy with high mortality and poor prognosis due to a lack of predictive markers. Increasing evidence has demonstrated small nucleolar RNAs (snoRNAs) play an important role in tumorigenesis. The aim of this study was to identify a prognostic snoRNA signature of HNSCC. Survival-related snoRNAs were screened by Cox regression analysis (univariate, least absolute shrinkage and selection operator, and multivariate). The predictive value was validated in different subgroups. The biological functions were explored by coexpression analysis and gene set enrichment analysis (GSEA). One hundred and thirteen survival-related snoRNAs were identified, and a five-snoRNA signature predicted prognosis with high sensitivity and specificity. Furthermore, the signature was applicable to patients of different sexes, ages, stages, grades, and anatomic subdivisions. Coexpression analysis and GSEA revealed the five-snoRNA are involved in regulating malignant phenotype and DNA/RNA editing. This five-snoRNA signature is not only a promising predictor of prognosis and survival but also a potential biomarker for patient stratification management.

Long noncoding RNA LEF1-AS1 binds with HNRNPL to boost the proliferation, migration, and invasion in osteosarcoma by enhancing the mRNA stability of LEF1

Osteosarcoma (OS) is the most frequent type of cancer that starts in the bones, with a rather high tendency to metastasize to other bones at the early stages. Although many types of research have demonstrated that long noncoding RNAs commonly take part in the development of various cancers, the modulating mechanism of LEF1-AS1 in OS was unknown yet. In this study, our results disclosed that LEF1-AS1, as well as LEF1, had higher expression levels in OS cells than that in normal bone cells. LEF1-AS1 knockdown dramatically inhibited the proliferation, migration, as well as invasion in OS, which proved that LEF1-AS1 contributed to the growth of OS. Furthermore, HNRNPL knockdown suppressed the expression of LEF1. LEF1-AS1 was confirmed to sponge HNRNPL and HNRNPL could bind with LEF1. Both LEF1-AS1 and HNRNPL could enhance the stability of LEF1 mRNA. LEF1-AS1 acted as a promoter in stimulating the Wnt signaling pathway in OS. In rescue experiments, overexpression of LEF1 partially offset the inhibition LEF1-AS1 knockdown brought in the proliferation, migration as well as invasion of OS cells. Collectively, this study had investigated that LEF1-AS1 bound with HNRNPL to promote OS cell proliferation, migration as well as invasion by enhancing the messenger RNA stability of LEF1.

LncRNA-SRA1 Suppresses Osteosarcoma Cell Proliferation While Promoting Cell Apoptosis

Objective:

Osteosarcoma is a common malignant bone tumor that is frequently found in the long bones of children and adolescents. The aim of this study is to examine long noncoding RNA-steroid receptor RNA activator 1 expression in osteosarcoma to explore the biological function of long noncoding RNA steroid receptor RNA activator 1 on proliferation, migration, and invasion along with apoptosis and its regulatory mechanism, which would facilitate the early diagnosis and targeted therapy of osteosarcoma.

Methods:

First, microarray analysis was applied to determine the expression of long noncoding RNAs in osteosarcoma tissues and paired normal tissues. Then, quantitative real-time polymerase chain reaction was utilized to validate microarray findings. Next, osteosarcoma cancerous cell lines SJSA-1 and U2OS were transfected with pcDNA3.1-SRA1 or pCMV-sh-SRA1 to increase or decrease steroid receptor RNA activator 1 expression levels, and microRNA-208a inhibitors, mimic to investigate the effects of microRNA-208a on osteosarcoma as well as the regulatory relation between long noncoding RNA steroid receptor RNA activator 1 and microRNA-208a. Cell proliferation was evaluated through Cell Counting Kit-8 and colony formation assays. Flow cytometry analysis was conducted to evaluate the apoptosis ratio. The migration and invasion abilities were measured using wound-healing and transwell assays.

Results:

Long noncoding RNA-steroid receptor RNA activator 1 expression was downregulated in osteosarcoma tissues and cells compared with that in corresponding normal tissues, whereas microRNA-208a expression was upregulated in osteosarcoma tissues. Moreover, the restoration of long noncoding RNA steroid receptor RNA activator 1 inhibited cell proliferation, and upregulation of long noncoding RNA steroid receptor RNA activator 1 restrained cell migration and invasion but boosted the apoptosis rate in osteosarcoma cells. In addition, long noncoding RNA steroid receptor RNA activator 1 targeting microRNA-208a was involved in the progression of osteosarcoma. Furthermore, upregulating microRNA-208a exerted similar roles of silencing long noncoding RNA steroid receptor RNA activator 1 in cell apoptosis, proliferation, migration, and invasion, which were reversed by enhancing the expression of long noncoding RNA steroid receptor RNA activator 1.

Conclusions:

In our study, long noncoding RNA steroid receptor RNA activator 1 played an antitumor role in osteosarcoma as it reduced cell migration, invasion, and proliferation, but facilitated cell apoptosis via sponging microRNA-208a, which could be regarded as a potential therapeutic target of osteosarcoma treatment.

Extracellular Nanovesicles Secreted by Human Osteosarcoma Cells Promote Angiogenesis

Angiogenesis involves a number of different players among which extracellular nanovesicles (EVs) have recently been proposed as an efficient cargo of pro-angiogenic mediators. Angiogenesis plays a key role in osteosarcoma (OS) development and progression. Acidity is a hallmark of malignancy in a variety of cancers, including sarcomas, as a result of an increased energetic metabolism. The aim of this study was to investigate the role of EVs derived from osteosarcoma cells on angiogenesis and whether extracellular acidity, generated by tumor metabolism, could influence EVs activity. For this purpose, we purified and characterized EVs from OS cells maintained at either acidic or neutral pH. The ability of EVs to induce angiogenesis was assessed in vitro by endothelial cell tube formation and in vivo using chicken chorioallantoic membrane. Our findings demonstrated that EVs derived from osteosarcoma cells maintained either in acidic or neutral conditions induced angiogenesis. The results showed that miRNA and protein content of EVs cargo are correlated with pro-angiogenic activity and this activity is increased by the acidity of tumor microenvironment. This study provides evidence that EVs released by human osteosarcoma cells act as carriers of active angiogenic stimuli that are able to promote endothelial cell functions relevant to angiogenesis.

A Four-Pseudogene Classifier Identified by Machine Learning Serves as a Novel Prognostic Marker for Survival of Osteosarcoma

Osteosarcoma is a common malignancy with high mortality and poor prognosis due to lack of predictive markers. Increasing evidence has demonstrated that pseudogenes, a type of non-coding gene, play an important role in tumorigenesis. The aim of this study was to identify a prognostic pseudogene signature of osteosarcoma by machine learning. A sample of 94 osteosarcoma patients’ RNA-Seq data with clinical follow-up information was involved in the study. The survival-related pseudogenes were screened and related signature model was constructed by cox-regression analysis (univariate, lasso, and multivariate). The predictive value of the signature was further validated in different subgroups. The putative biological functions were determined by co-expression analysis. In total, 125 survival-related pseudogenes were identified and a four-pseudogene (RPL11-551L14.1, HR: 0.65 (95% CI: 0.44–0.95); RPL7AP28, HR: 0.32 (95% CI: 0.14–0.76); RP4-706A16.3, HR: 1.89 (95% CI: 1.35–2.65); RP11-326A19.5, HR: 0.52(95% CI: 0.37–0.74)) signature effectively distinguished the high- and low-risk patients, and predicted prognosis with high sensitivity and specificity (AUC: 0.878). Furthermore, the signature was applicable to patients of different genders, ages, and metastatic status. Co-expression analysis revealed the four pseudogenes are involved in regulating malignant phenotype, immune, and DNA/RNA editing. This four-pseudogene signature is not only a promising predictor of prognosis and survival, but also a potential marker for monitoring therapeutic schedule. Therefore, our findings may have potential clinical significance.