Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls
Personalized oncology is a rapidly evolving area and offers cancer patients therapy options that are more specific than ever. However, there is still a lack of understanding regarding transcriptomic similarities or differences of metastases and corresponding primary sites. Applying two unsupervised...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-07-01
|
| Series: | Genes |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4425/13/8/1335 |
| _version_ | 1797445232712744960 |
|---|---|
| author | André Marquardt Philip Kollmannsberger Markus Krebs Antonella Argentiero Markus Knott Antonio Giovanni Solimando Alexander Georg Kerscher |
| author_facet | André Marquardt Philip Kollmannsberger Markus Krebs Antonella Argentiero Markus Knott Antonio Giovanni Solimando Alexander Georg Kerscher |
| author_sort | André Marquardt |
| collection | DOAJ |
| description | Personalized oncology is a rapidly evolving area and offers cancer patients therapy options that are more specific than ever. However, there is still a lack of understanding regarding transcriptomic similarities or differences of metastases and corresponding primary sites. Applying two unsupervised dimension reduction methods (t-Distributed Stochastic Neighbor Embedding (t-SNE) and Uniform Manifold Approximation and Projection (UMAP)) on three datasets of metastases (<i>n</i> = 682 samples) with three different data transformations (unprocessed, log10 as well as log10 + 1 transformed values), we visualized potential underlying clusters. Additionally, we analyzed two datasets (<i>n</i> = 616 samples) containing metastases and primary tumors of one entity, to point out potential familiarities. Using these methods, no tight link between the site of resection and cluster formation outcome could be demonstrated, or for datasets consisting of solely metastasis or mixed datasets. Instead, dimension reduction methods and data transformation significantly impacted visual clustering results. Our findings strongly suggest data transformation to be considered as another key element in the interpretation of visual clustering approaches along with initialization and different parameters. Furthermore, the results highlight the need for a more thorough examination of parameters used in the analysis of clusters. |
| first_indexed | 2024-03-09T13:23:47Z |
| format | Article |
| id | doaj.art-d7c80cbebf3b4eb38fb7ad1df266eeb9 |
| institution | Directory Open Access Journal |
| issn | 2073-4425 |
| language | English |
| last_indexed | 2024-03-09T13:23:47Z |
| publishDate | 2022-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Genes |
| spelling | doaj.art-d7c80cbebf3b4eb38fb7ad1df266eeb92023-11-30T21:27:21ZengMDPI AGGenes2073-44252022-07-01138133510.3390/genes13081335Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel PitfallsAndré Marquardt0Philip Kollmannsberger1Markus Krebs2Antonella Argentiero3Markus Knott4Antonio Giovanni Solimando5Alexander Georg Kerscher6Institute of Pathology, Klinikum Stuttgart, 70174 Stuttgart, GermanyCenter for Computational and Theoretical Biology, University of Würzburg, 97074 Würzburg, GermanyComprehensive Cancer Center Mainfranken, University Hospital Würzburg, 97080 Würzburg, GermanyIRCCS Istituto Tumori “Giovanni Paolo II” of Bari, 70124 Bari, ItalyDepartment of Hematology, Oncology, Stem Cell Transplantation and Palliative Care, Klinikum Stuttgart, 70174 Stuttgart, GermanyIRCCS Istituto Tumori “Giovanni Paolo II” of Bari, 70124 Bari, ItalyComprehensive Cancer Center Mainfranken, University Hospital Würzburg, 97080 Würzburg, GermanyPersonalized oncology is a rapidly evolving area and offers cancer patients therapy options that are more specific than ever. However, there is still a lack of understanding regarding transcriptomic similarities or differences of metastases and corresponding primary sites. Applying two unsupervised dimension reduction methods (t-Distributed Stochastic Neighbor Embedding (t-SNE) and Uniform Manifold Approximation and Projection (UMAP)) on three datasets of metastases (<i>n</i> = 682 samples) with three different data transformations (unprocessed, log10 as well as log10 + 1 transformed values), we visualized potential underlying clusters. Additionally, we analyzed two datasets (<i>n</i> = 616 samples) containing metastases and primary tumors of one entity, to point out potential familiarities. Using these methods, no tight link between the site of resection and cluster formation outcome could be demonstrated, or for datasets consisting of solely metastasis or mixed datasets. Instead, dimension reduction methods and data transformation significantly impacted visual clustering results. Our findings strongly suggest data transformation to be considered as another key element in the interpretation of visual clustering approaches along with initialization and different parameters. Furthermore, the results highlight the need for a more thorough examination of parameters used in the analysis of clusters.https://www.mdpi.com/2073-4425/13/8/1335visual clusteringt-SNEUMAPtranscriptomic analysiscancermetastasis |
| spellingShingle | André Marquardt Philip Kollmannsberger Markus Krebs Antonella Argentiero Markus Knott Antonio Giovanni Solimando Alexander Georg Kerscher Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls Genes visual clustering t-SNE UMAP transcriptomic analysis cancer metastasis |
| title | Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls |
| title_full | Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls |
| title_fullStr | Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls |
| title_full_unstemmed | Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls |
| title_short | Visual Clustering of Transcriptomic Data from Primary and Metastatic Tumors—Dependencies and Novel Pitfalls |
| title_sort | visual clustering of transcriptomic data from primary and metastatic tumors dependencies and novel pitfalls |
| topic | visual clustering t-SNE UMAP transcriptomic analysis cancer metastasis |
| url | https://www.mdpi.com/2073-4425/13/8/1335 |
| work_keys_str_mv | AT andremarquardt visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls AT philipkollmannsberger visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls AT markuskrebs visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls AT antonellaargentiero visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls AT markusknott visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls AT antoniogiovannisolimando visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls AT alexandergeorgkerscher visualclusteringoftranscriptomicdatafromprimaryandmetastatictumorsdependenciesandnovelpitfalls |