- Predictors of breast cancer cell types and their prognostic power in breast cancer patients. - We investigate if the “ cell type ” of a cancer cell can be predicted based on the expression profiles of a small set of transcripts.. - Results: We outline a predictive analytics pipeline to accurately predict 6 breast cancer cell types using single cell gene expression profiles. - A signature risk score originating from 65 protein coding genes and 5 lncRNA predictors is associated with prognostic survival of TCGA breast cancer patients. - We further show that predictors restricted to a particular cell type serve as better prognostic markers for the respective patient subtype.. - Conclusion: Our results show that in general, the breast cancer cell type predictors are also associated with patient survivability and hence have clinical significance.. - 1%) muta- tions that may play an important role in tumor evolution and therapy resistance were identified in breast cancer [9]. - In another study, scRNASeq was used to study the spread of single circulating tumor cells and circulat- ing tumor cell clusters in metastatic breast cancer patients and mouse models [13].. - Since most of the publicly available data use few numbers of cells, we focused on a breast cancer study that used an ample number of cells originating from 6 different cell types. - Our study indicates that the cell type of a breast cancer cell can be expressed as dependent variable of expression profiles of a small set of transcripts. - risk score using 68 protein coding genes and 5 lncRNAs predictors is associated with prognostic survival of TCGA breast cancer patients. - We further show that predictors restricted to a particular cell type serve as better prognostic markers for the respective patient subtype. - This shows that in general, the breast cancer cell type predictors are also associated with patient survivability and hence have clinical significance.. - Breast cancer single cell transcriptome data. - However, we focused only on one scRNASeq data on breast cancer cells because this study used ample number of cells, the raw data was publicly available, and cell type information for predictive analytics was also available. - The scRNASeq breast cancer data had 401 cells and originated from 6 different cell types populations: 15% estrogen receptor positive (ER. - 29% triple-negative breast cancer (TNBC), and 8% BC07LN (lymph node metastasized triple-negative).. - The second part of the analysis involves tuning various pre- dictive models for accurate prediction of cell type of breast cancer cells.. - These steps reduced the number of genes in breast cancer data set to 5592 genes. - To make sure that there was no bias in the split for a particular cell type;. - machines (SVM) showed the best predictive ability of the breast cancer cell types. - a Panel (a) show the steps involved in removing the number of genes excluded from the set of predictors for breast cancer cell type. - These results indicate that when considered separ- ately, lncRNAs and Protein-coding genes are weaker pre- dictors for breast cancer cell types. - However, the right mixture of protein-coding genes and lncRNAs act serve as accurate predictors for breast cancer cell types. - Interestingly 6 clusters showed a restricted expression pattern to one single cell type (Fig. - These results show that expression of the majority of the predictors is restricted to a single breast cancer cell type.. - We found that some known breast cancer markers were included in our 308-predictor set. - cell type (Fig. - ERBB2 gene, an important marker for HER2+ cancer patients, is also included in our predictor set and its expression is restricted to HER2 + breast cancer cell type (Fig. - breast cancer type [20] is one of the predictor belong- ing to cluster G (Fig. - Hsp90 proteins overexpression has been proposed to have some role in making breast cancer cells become resistant to various stress stimuli [22. - Note that the 308 predictors can dis- criminate the 6-different breast cancer cell types.. - particular breast cancer cell type may not be overall good candidate for differentiating between the 6 can- cer cell types. - Survival analysis of TCGA cancer patients using cell type predictors. - Next, we checked if the predictor genes have prognostic power in stratifying breast cancer patient risk and likeli- hood of survival. - We gathered clinical data for 816 breast cancer patients from the cancer genome atlas project (TCGA). - These results suggest that the presence of alteration in top breast cancer cell type pre- dictors can be an indicator for likelihood of survival in breast invasive carcinoma patients.. - Next, we checked if the expression signature of the predictors could be used for survival prognosis of TCGA breast cancer patients. - The TCGA breast cancer patients were partitioned into training (80%) and testing (20%) data sets. - Using the training data set, we identified indi- vidual cell type predictors with a prognostic power for the TCGA breast cancer patients. - In addition, the 73 predictors’ signature was further applied to the entire TCGA breast cancer dataset. - 4 Survival curve analysis of TCGA breast cancer patients based on cancer cell type predictors. - Cell type specific predictors serve as better prognostic markers for respective patient subtype. - 5 and Additional file 1: Figure S1, we showed that the cell type predictors can serve as prognostic markers in breast cancer patients irrespective of the cancer sub- type the patient belongs to. - 3, which were specific to different types of breast cancer cells. - We explored if the predictors restricted to a particular cell type showed better prognostic significance for the respective breast cancer subtype in TCGA patients. - 5 Prognostic power of breast cancer cell type predictors. - The TCGA breast cancer patients were stratified into high risk and low risk patients based on the expression profiles of the breast cancer cell type predictors in TCGA patients. - We found that predictors restricted to a par- ticular cell type showed better prognostic significance for the respective breast cancer subtype in TCGA pa- tients (Fig. - Cell type specific predictors serve as prognostic markers for early stage breast cancers. - Providing good prognosis to early stage breast cancer patients is crucial. - Therefore, we further checked if the cell type specific predictors can also serve as good prog- nostic markers in early stage breast cancer patients.. - They were analyzed separately using a multivariate Cox proportional hazard regression model only on the TCGA patients belonging to the subtype representing the cell type of interest. - Similarly, HER2+ and TNBC cell type specific predictors serve as better prognostic markers in HER2+ and TNBC patients respectively. - showed that cell type specific predictors showed no significant difference in the expression levels across the stages (Fig. - We found that each cell specific predictors served as good prognostic markers for early stage breast cancer patients (Fig. - Our results indicate that the predictors for breast cancer patients can also serve as prognostic markers for early stage breast cancer patients.. - HER2-enriched, and Basal-like (TNBC) subtypes of breast cancer.. - In this study, we have successfully outlined a predictive analytics pipeline to accurately predict 6 breast cancer cell types using single cell gene expression profiles. - 7 Prognostic potential of predictors in early stage breast cancer patients. - a Expression of predictors specific to ER+, HER2+, and TNBC cell type predictors in cells at different stages. - b Prognostic potential of predictors specific to ER+, HER2+, and TNBC types in stage I TCGA breast cancer patients. - Additionally, we also highlight 6 clusters of genes that accordingly restricted to 6 different breast cancer cell types and validated the results by demonstrating several previously established breast cancer markers in each cluster. - A signature risk score originating from 65 pro- tein coding genes and 5 lncRNA predictors is associated with prognostic survival of TCGA breast cancer patients.. - Therefore, a subset of the breast cancer cell type predic- tors is also associated with patient survivability and hence have clinical significance.. - This indicates that the cell type. - of a breast cancer cell can be expressed as dependent variable of expression profiles of a set of predictors;. - We were not able to find a different independ- ent single cell sequencing data for breast cancer with known cell types to further validate our predictive model. - Recently, lncRNAs are also emerging as prognostic markers in different cancer including breast cancer. - For the first time, our study reports lncRNAs which can discriminate different breast cancer cell types and also have prognostic potential. - In our study, we found that the signature score of a set of 5 lncRNAs (ENSG ENSG ENSG ENSG and ENSG can be used for breast cancer prog- nosis. - Signature risk scores of sets of 12 lncRNAs [24], 9 lncRNAs [25], and 4 lncRNAs [26] have been reported by independent studies to have potential prognostic power in breast cancer patients. - This shows that this set of 5 lncRNAs contrib- ute to the discrimination of different breast cancer cell type as well have prognostic significance in breast cancer patients. - Specific to breast cancer, there is a study that explored the oncogenic landscape of lncRNAs in breast cancer patients [27]. - They also reported that lncRNAs, HOTAIR, LINC00115, MCM3AP-AS1, TINCR, PPP1R26-AS1, and DSCAM-AS1 were breast cancer prognosis-associated lncRNAs. - This study simply uses stat- istical methods of differential gene expression analysis to identify dysregulated lncRNAs in different subtypes of breast cancer. - We also employed machine learning models and clustering to val- idate that these genes and lncRNAs are not only good pre- dictors of breast cancer subtypes but also can be grouped into subtype specific predictors. - This shows that studying breast cancer using different methods can complement each other and are necessary in deciphering the under- lying regulatory layers.. - In summary, this study both validates the use of scRNA-seq to transcriptionally profile an ample number of cells originating from 6 different breast cancer cell types and defines 65 protein coding genes and 5 lncRNAs which are significantly related to prognostic survival of breast cancer patients from TCGA database.. - Thus, we must reinforce the need for longitudinal characterizations of tumor transcriptomes by using our analytic pipeline to predict predominant breast cancer cell types so as to guide more personalized clinical care.. - Here, we outline a predictive analytics pipeline to accurately predict 6 breast cancer cell types using single cell gene expression profiles. - Using machine learning techniques, we identify 308 predictors, out of which 34 are long non-coding RNAs, of breast cancer cell types.. - This set of predictor are able to identify different breast cancer cell types with 98% prediction accuracies. - We further show that a signature risk score originating from 65 protein coding genes and 5 lncRNA predictors is associated with prognostic survival of TCGA breast cancer patients. - We further show that predictors restricted to a particular cell type serve as better prog- nostic markers for the respective patient subtype. - Our results show that in general, the breast cancer cell type predictors are also associated with patient survivability and hence have clinical significance.. - To make sure that there was no bias in the split for a particular cell type. - The univariate Cox regression analysis was performed to examine the relationship between the expression levels of breast cancer cell type predictors in TCGA breast cancer patients and the overall survivability from the training set with an aim to determine which predictors could potentially be of functional significance in breast cancer prognosis. - Prognostic potential of 6 sets of cell specific predictors in TCGA breast cancer patients, Cells belonging to the same type cluster together. - Clonal evolution in breast cancer revealed by single nucleus genome sequencing. - Circulating tumor cell clusters are oligoclonal precursors of breast cancer metastasis. - Single cell profiling of circulating tumor cells: transcriptional heterogeneity and diversity from breast cancer cell lines. - Gene expression profiling of breast cancer cell lines in response to soy isoflavones using a pangenomic microarray approach. - Role of the androgen receptor in triple-negative breast cancer. - The clinical significance of androgen receptors in breast cancer and their relation to histological and cell biological parameters. - Heat shock protein 90 (Hsp90) expression and breast cancer. - Discovery of potential prognostic long non-coding RNA biomarkers for predicting the risk of tumor recurrence of breast cancer patients. - A potential prognostic long non-coding RNA signature to predict metastasis-free survival of breast cancer patients. - A four-long non-coding RNA signature in predicting breast cancer survival. - Oncogenic long noncoding RNA landscape in breast cancer
Xem thử không khả dụng, vui lòng xem tại trang nguồn hoặc xem
Tóm tắt