Cargando…
scROSHI: robust supervised hierarchical identification of single cells
Identifying cell types based on expression profiles is a pillar of single cell analysis. Existing machine-learning methods identify predictive features from annotated training data, which are often not available in early-stage studies. This can lead to overfitting and inferior performance when appli...
Autores principales: | , , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Oxford University Press
2023
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10273189/ https://www.ncbi.nlm.nih.gov/pubmed/37332656 http://dx.doi.org/10.1093/nargab/lqad058 |
_version_ | 1785059645536075776 |
---|---|
author | Prummer, Michael Bertolini, Anne Bosshard, Lars Barkmann, Florian Yates, Josephine Boeva, Valentina Stekhoven, Daniel Singer, Franziska |
author_facet | Prummer, Michael Bertolini, Anne Bosshard, Lars Barkmann, Florian Yates, Josephine Boeva, Valentina Stekhoven, Daniel Singer, Franziska |
author_sort | Prummer, Michael |
collection | PubMed |
description | Identifying cell types based on expression profiles is a pillar of single cell analysis. Existing machine-learning methods identify predictive features from annotated training data, which are often not available in early-stage studies. This can lead to overfitting and inferior performance when applied to new data. To address these challenges we present scROSHI, which utilizes previously obtained cell type-specific gene lists and does not require training or the existence of annotated data. By respecting the hierarchical nature of cell type relationships and assigning cells consecutively to more specialized identities, excellent prediction performance is achieved. In a benchmark based on publicly available PBMC data sets, scROSHI outperforms competing methods when training data are limited or the diversity between experiments is large. |
format | Online Article Text |
id | pubmed-10273189 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2023 |
publisher | Oxford University Press |
record_format | MEDLINE/PubMed |
spelling | pubmed-102731892023-06-17 scROSHI: robust supervised hierarchical identification of single cells Prummer, Michael Bertolini, Anne Bosshard, Lars Barkmann, Florian Yates, Josephine Boeva, Valentina Stekhoven, Daniel Singer, Franziska NAR Genom Bioinform Methods Article Identifying cell types based on expression profiles is a pillar of single cell analysis. Existing machine-learning methods identify predictive features from annotated training data, which are often not available in early-stage studies. This can lead to overfitting and inferior performance when applied to new data. To address these challenges we present scROSHI, which utilizes previously obtained cell type-specific gene lists and does not require training or the existence of annotated data. By respecting the hierarchical nature of cell type relationships and assigning cells consecutively to more specialized identities, excellent prediction performance is achieved. In a benchmark based on publicly available PBMC data sets, scROSHI outperforms competing methods when training data are limited or the diversity between experiments is large. Oxford University Press 2023-06-16 /pmc/articles/PMC10273189/ /pubmed/37332656 http://dx.doi.org/10.1093/nargab/lqad058 Text en © The Author(s) 2023. Published by Oxford University Press on behalf of NAR Genomics and Bioinformatics. https://creativecommons.org/licenses/by/4.0/This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. |
spellingShingle | Methods Article Prummer, Michael Bertolini, Anne Bosshard, Lars Barkmann, Florian Yates, Josephine Boeva, Valentina Stekhoven, Daniel Singer, Franziska scROSHI: robust supervised hierarchical identification of single cells |
title | scROSHI: robust supervised hierarchical identification of single cells |
title_full | scROSHI: robust supervised hierarchical identification of single cells |
title_fullStr | scROSHI: robust supervised hierarchical identification of single cells |
title_full_unstemmed | scROSHI: robust supervised hierarchical identification of single cells |
title_short | scROSHI: robust supervised hierarchical identification of single cells |
title_sort | scroshi: robust supervised hierarchical identification of single cells |
topic | Methods Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10273189/ https://www.ncbi.nlm.nih.gov/pubmed/37332656 http://dx.doi.org/10.1093/nargab/lqad058 |
work_keys_str_mv | AT prummermichael scroshirobustsupervisedhierarchicalidentificationofsinglecells AT bertolinianne scroshirobustsupervisedhierarchicalidentificationofsinglecells AT bosshardlars scroshirobustsupervisedhierarchicalidentificationofsinglecells AT barkmannflorian scroshirobustsupervisedhierarchicalidentificationofsinglecells AT yatesjosephine scroshirobustsupervisedhierarchicalidentificationofsinglecells AT boevavalentina scroshirobustsupervisedhierarchicalidentificationofsinglecells AT scroshirobustsupervisedhierarchicalidentificationofsinglecells AT stekhovendaniel scroshirobustsupervisedhierarchicalidentificationofsinglecells AT singerfranziska scroshirobustsupervisedhierarchicalidentificationofsinglecells |