Overcoming limitations in sparse cell-type predictions for a biologically reliable task.
<p><b>A</b>, Illustration of mouse hippocampus on mouse brain coronal section. <b>B</b>, 2D UMAP representation of the 8 HPF cell types. <b>C</b>, Overlapping CA1 and CA3 due to cell type similarity, with cluster radii calculated using centroid distances and...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | |
| منشور في: |
2025
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| _version_ | 1852019349717516288 |
|---|---|
| author | Nuray Sogunmez Erdogan (21533872) |
| author2 | Deniz Eroglu (21533875) |
| author2_role | author |
| author_facet | Nuray Sogunmez Erdogan (21533872) Deniz Eroglu (21533875) |
| author_role | author |
| dc.creator.none.fl_str_mv | Nuray Sogunmez Erdogan (21533872) Deniz Eroglu (21533875) |
| dc.date.none.fl_str_mv | 2025-06-12T18:57:39Z |
| dc.identifier.none.fl_str_mv | 10.1371/journal.pcbi.1013169.g002 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/figure/Overcoming_limitations_in_sparse_cell-type_predictions_for_a_biologically_reliable_task_/29309908 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Genetics Biotechnology Cancer Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified unlike conventional approaches understanding cellular responses resolution cellular maps preserving biological coherence predicted cellular landscapes machine learning algorithm e ., spatially decode tissue organization induced sparse regression functionally consistent cell distinct cell types leveraging sparse cell integrating reference single cell type medleys spatial transcriptomics spot spatial transcriptomics sparse deconvolution integrates spot cell data spatial datasets type variations type localization type distributions type arrangements wispr ), unmatched datasets specific hyperparameters signaling however practical utility permissive algorithms neglecting biology neglect biology methods rely introduce weight inaccurate predictions grounded constraints genome coverage driven modeling cancerous tissues |
| dc.title.none.fl_str_mv | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| dc.type.none.fl_str_mv | Image Figure info:eu-repo/semantics/publishedVersion image |
| description | <p><b>A</b>, Illustration of mouse hippocampus on mouse brain coronal section. <b>B</b>, 2D UMAP representation of the 8 HPF cell types. <b>C</b>, Overlapping CA1 and CA3 due to cell type similarity, with cluster radii calculated using centroid distances and cell locations at maximum distances. <b>D-I</b>, Model predictions for CA3 in the P8 and <b>J-O</b> in adult mouse brain are presented. <b>D, J</b>, In DWLS prediction, CA3 cells are slightly over-represented in mouse brain spots (blue shade), with frequent spot prediction failures (red). <b>E, K</b>, RCTD prediction reveals high percentages of cells in spots located in hippocampal, thalamic and cortical regions. <b>F, L</b>, S-DWLS prediction displays confined but still disproportionate of cells in spots concentrated in stratum oriens, stratum pyramidale and stratum lucidum and thalamus regions. <b>G, M</b>, SPOTlight predicts a high abundance of CA3 cells in DG and across various brain zones including cortex. <b>H, N</b>, Stereoscope predicts CA3 in almost all hippocampus, thalamus and cortical regions, while <b>I, O</b> WISpR predicts CA3 cells in exactly CA3 morphological region in both P8 and adult mice, supporting its more accurate prediction on mismatched datasets.</p> |
| eu_rights_str_mv | openAccess |
| id | Manara_ea7c0e90fbf5aaa653c0b0bb7a038ec6 |
| identifier_str_mv | 10.1371/journal.pcbi.1013169.g002 |
| network_acronym_str | Manara |
| network_name_str | ManaraRepo |
| oai_identifier_str | oai:figshare.com:article/29309908 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | Overcoming limitations in sparse cell-type predictions for a biologically reliable task.Nuray Sogunmez Erdogan (21533872)Deniz Eroglu (21533875)GeneticsBiotechnologyCancerBiological Sciences not elsewhere classifiedMathematical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedunlike conventional approachesunderstanding cellular responsesresolution cellular mapspreserving biological coherencepredicted cellular landscapesmachine learning algorithme ., spatiallydecode tissue organizationinduced sparse regressionfunctionally consistent celldistinct cell typesleveraging sparse cellintegrating reference singlecell type medleysspatial transcriptomics spotspatial transcriptomicssparse deconvolutionintegrates spotcell dataspatial datasetstype variationstype localizationtype distributionstype arrangementswispr ),unmatched datasetsspecific hyperparameterssignaling howeverpractical utilitypermissive algorithmsneglecting biologyneglect biologymethods relyintroduce weightinaccurate predictionsgrounded constraintsgenome coveragedriven modelingcancerous tissues<p><b>A</b>, Illustration of mouse hippocampus on mouse brain coronal section. <b>B</b>, 2D UMAP representation of the 8 HPF cell types. <b>C</b>, Overlapping CA1 and CA3 due to cell type similarity, with cluster radii calculated using centroid distances and cell locations at maximum distances. <b>D-I</b>, Model predictions for CA3 in the P8 and <b>J-O</b> in adult mouse brain are presented. <b>D, J</b>, In DWLS prediction, CA3 cells are slightly over-represented in mouse brain spots (blue shade), with frequent spot prediction failures (red). <b>E, K</b>, RCTD prediction reveals high percentages of cells in spots located in hippocampal, thalamic and cortical regions. <b>F, L</b>, S-DWLS prediction displays confined but still disproportionate of cells in spots concentrated in stratum oriens, stratum pyramidale and stratum lucidum and thalamus regions. <b>G, M</b>, SPOTlight predicts a high abundance of CA3 cells in DG and across various brain zones including cortex. <b>H, N</b>, Stereoscope predicts CA3 in almost all hippocampus, thalamus and cortical regions, while <b>I, O</b> WISpR predicts CA3 cells in exactly CA3 morphological region in both P8 and adult mice, supporting its more accurate prediction on mismatched datasets.</p>2025-06-12T18:57:39ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pcbi.1013169.g002https://figshare.com/articles/figure/Overcoming_limitations_in_sparse_cell-type_predictions_for_a_biologically_reliable_task_/29309908CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/293099082025-06-12T18:57:39Z |
| spellingShingle | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. Nuray Sogunmez Erdogan (21533872) Genetics Biotechnology Cancer Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified unlike conventional approaches understanding cellular responses resolution cellular maps preserving biological coherence predicted cellular landscapes machine learning algorithm e ., spatially decode tissue organization induced sparse regression functionally consistent cell distinct cell types leveraging sparse cell integrating reference single cell type medleys spatial transcriptomics spot spatial transcriptomics sparse deconvolution integrates spot cell data spatial datasets type variations type localization type distributions type arrangements wispr ), unmatched datasets specific hyperparameters signaling however practical utility permissive algorithms neglecting biology neglect biology methods rely introduce weight inaccurate predictions grounded constraints genome coverage driven modeling cancerous tissues |
| status_str | publishedVersion |
| title | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| title_full | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| title_fullStr | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| title_full_unstemmed | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| title_short | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| title_sort | Overcoming limitations in sparse cell-type predictions for a biologically reliable task. |
| topic | Genetics Biotechnology Cancer Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified unlike conventional approaches understanding cellular responses resolution cellular maps preserving biological coherence predicted cellular landscapes machine learning algorithm e ., spatially decode tissue organization induced sparse regression functionally consistent cell distinct cell types leveraging sparse cell integrating reference single cell type medleys spatial transcriptomics spot spatial transcriptomics sparse deconvolution integrates spot cell data spatial datasets type variations type localization type distributions type arrangements wispr ), unmatched datasets specific hyperparameters signaling however practical utility permissive algorithms neglecting biology neglect biology methods rely introduce weight inaccurate predictions grounded constraints genome coverage driven modeling cancerous tissues |