Classifying Medulloblastoma Subgroups Based on Small, Clinically Achievable Gene Sets

Gershanov, Sivan; Madiwale, Shreyas; Feinberg-Gorenshtein, Galina; Vainer, Igor; Nehushtan, Tamar; Michowiz, Shalom; Goldenberg-Cohen, Nitza; Birger, Yehudit; Toledano, Helen; Salmon‐Divon, Mali

doi:10.3389/fonc.2021.637482

Cited by 8 publications

(5 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Risk stratification and accurate outcome prediction of future Grp 3 MB cohorts in the absence of high-throughput profiling techniques may be enhanced by assessing KIRREL2 expression in routine neuropathology. For example, single gene RQ-PCR quantification, Taqman low-density arrays, or Nanostring-based analyses evaluating expression of this gene might be easily developed in neuropathological practice after elaboration of optimal cut-off levels for each method applied [ 1 , 5 , 6 , 10 , 14 , 31 ]. In addition, KIRREL2 protein expression was defined here as a prognostic indicator.…”

Section: Discussionmentioning

confidence: 99%

Gene expression profiling of Group 3 medulloblastomas defines a clinically tractable stratification based on KIRREL2 expression

et al. 2022

View full text Add to dashboard Cite

Medulloblastomas (MB) molecularly designated as Group 3 (Grp 3) MB represent a more clinically aggressive tumor variant which, as a group, displays heterogeneous molecular characteristics and disease outcomes. Reliable risk stratification of Grp 3 MB would allow for appropriate assignment of patients to aggressive treatment protocols and, vice versa, for sparing adverse effects of high-dose radio-chemotherapy in patients with standard or low-risk tumors. Here we performed RNA-based analysis on an international cohort of 179 molecularly designated Grp 3 MB treated with HIT protocols. We analyzed the clinical significance of differentially expressed genes, thereby developing optimal prognostic subdivision of this MB molecular group. We compared the transcriptome profiles of two Grp 3 MB subsets with various outcomes (76 died within the first 60 months vs. 103 survived this period) and identified 224 differentially expressed genes (DEG) between these two clinical groups (Limma R algorithm, adjusted p-value < 0.05). We selected the top six DEG overexpressed in the unfavorable cohort for further survival analysis and found that expression of all six genes strongly correlated with poor outcomes. However, only high expression of KIRREL2 was identified as an independent molecular prognostic indicator of poor patients’ survival. Based on clinical and molecular patterns, four risk categories were outlined for Grp 3 MB patients: i. low-risk: M0-1/MYC non-amplified/KIRREL2 low (n = 48; 5-year OS—95%); ii. standard-risk: M0-1/MYC non-amplified/KIRREL2 high or M2-3/MYC non-amplified/KIRREL2 low (n = 65; 5-year OS—70%); iii. high-risk: M2-3/MYC non-amplified/KIRREL2 high (n = 36; 5-year OS—30%); iv. very high risk—all MYC amplified tumors (n = 30; 5-year OS—0%). Cross-validated survival models incorporating KIRREL2 expression with clinical features allowed for the reclassification of up to 50% of Grp 3 MB patients into a more appropriate risk category. Finally, KIRREL2 immunopositivity was also identified as a predictive indicator of Grp 3 MB poor survival, thus suggesting its application as a possible prognostic marker in routine clinical settings. Our results indicate that integration of KIRREL2 expression in risk stratification models may improve Grp 3 MB outcome prediction. Therefore, simple gene and/or protein expression analyses for this molecular marker could be easily adopted for Grp 3 MB prognostication and may help in assigning patients to optimal therapeutic approaches in prospective clinical trials.

show abstract

Section: Discussionmentioning

confidence: 99%

Gene expression profiling of Group 3 medulloblastomas defines a clinically tractable stratification based on KIRREL2 expression

et al. 2022

View full text Add to dashboard Cite

show abstract

“…To illustrate the limitation of bulk simulation using randomly selected cells (namely homogeneous simulation), we considered a simulation setup using single cell Medulloblastoma (MB) dataset from Riemondy et al [ 31 ]. Medulloblastoma is a well-recognized heterogeneous brain cancer with four distinct subtypes based on genetic characteristics: WNT, SHH, Group 3, and Group 4 [ 48 , 49 ]. Analyzing the tSNE clustering of malignant populations (Fig.…”

Section: Resultsmentioning

confidence: 99%

Heterogeneous pseudobulk simulation enables realistic benchmarking of cell-type deconvolution methods

Hu,

Chikina

2024

Genome Biol

View full text Add to dashboard Cite

Background Computational cell type deconvolution enables the estimation of cell type abundance from bulk tissues and is important for understanding tissue microenviroment, especially in tumor tissues. With rapid development of deconvolution methods, many benchmarking studies have been published aiming for a comprehensive evaluation for these methods. Benchmarking studies rely on cell-type resolved single-cell RNA-seq data to create simulated pseudobulk datasets by adding individual cells-types in controlled proportions. Results In our work, we show that the standard application of this approach, which uses randomly selected single cells, regardless of the intrinsic difference between them, generates synthetic bulk expression values that lack appropriate biological variance. We demonstrate why and how the current bulk simulation pipeline with random cells is unrealistic and propose a heterogeneous simulation strategy as a solution. The heterogeneously simulated bulk samples match up with the variance observed in real bulk datasets and therefore provide concrete benefits for benchmarking in several ways. We demonstrate that conceptual classes of deconvolution methods differ dramatically in their robustness to heterogeneity with reference-free methods performing particularly poorly. For regression-based methods, the heterogeneous simulation provides an explicit framework to disentangle the contributions of reference construction and regression methods to performance. Finally, we perform an extensive benchmark of diverse methods across eight different datasets and find BayesPrism and a hybrid MuSiC/CIBERSORTx approach to be the top performers. Conclusions Our heterogeneous bulk simulation method and the entire benchmarking framework is implemented in a user friendly package https://github.com/humengying0907/deconvBenchmarking and https://doi.org/10.5281/zenodo.8206516, enabling further developments in deconvolution methods.

show abstract

“…Medulloblastoma is a well-recognized heterogeneous brain cancer with four distinct subtypes based on genetic characteristics. WNT, SHH, Group 3 and Group 4 [25, 26]. According to the tSNE clustering, malignant cells within the same molecular subtypes are largely clustered together, indicating underlying differences between subtypes.…”

Section: Resultsmentioning

confidence: 99%

Heterogeneous pseudobulk simulation enables realistic benchmarking of cell-type deconvolution methods

Chikina

2023

Preprint

View full text Add to dashboard Cite

Computational cell type deconvolution enables estimation of cell type abundance from bulk tissues and is important for understanding cell-cell interactions, especially in tumor tissues. With rapid development of deconvolution methods, many benchmarking studies have been published aiming for a comprehensive evaluation for these methods. Benchmarking studies rely on cell-type resolved single cell RNA-seq data to create simulated pseudbulk datasets by adding individual cells-types in controlled proportions. In our work we show that the standard application of this approach, which uses randomly selected single cells, regardless of the intrinsic difference between them, generates synthetic bulk expression values that lack appropriate biological variance. We demonstratewhy and how the current bulk simulation pipeline with random cells is unrealistic and propose a heterogeneous simulation strategy as a solution. Our heterogeneously simulated samples show realistic variance across hallmark gene-sets when comparing with real bulk samples from the TCGA dataset of the same tumor type. Using this new simulation pipeline to benchmark deconvolution methods we show that introducing biological heterogeneity has a notable effect on the results. Evaluating the robustness of different deconvolution approaches to heterogeneous simulation we find that reference-free methods that rely on simplex estimation perform poorly, marker-based methods and BayesPrism are most robust, while regress-based approaches fall in between. Importantly, we find that under the heterogeneous scenario marker based methods and BayesPrism outperform state of the art reference methods. Our findings highlight how different conceptual approaches can negate unmodeled heterogeneity and suggest that there is room for further methodological development.

show abstract

Classifying Medulloblastoma Subgroups Based on Small, Clinically Achievable Gene Sets

Cited by 8 publications

References 32 publications

Gene expression profiling of Group 3 medulloblastomas defines a clinically tractable stratification based on KIRREL2 expression

Gene expression profiling of Group 3 medulloblastomas defines a clinically tractable stratification based on KIRREL2 expression

Heterogeneous pseudobulk simulation enables realistic benchmarking of cell-type deconvolution methods

Heterogeneous pseudobulk simulation enables realistic benchmarking of cell-type deconvolution methods

Contact Info

Product

Resources

About