tetano
Editor, Senior Moderator
Nat Biotechnol
. 2022 Feb 28.
doi: 10.1038/s41587-021-01186-x. Online ahead of print.
Multiscale PHATE identifies multimodal signatures of COVID-19
Manik Kuchroo[SUP] #[/SUP][SUP] 1 [/SUP], Jessie Huang[SUP] #[/SUP][SUP] 2 [/SUP], Patrick Wong[SUP] #[/SUP][SUP] 3 [/SUP], Jean-Christophe Grenier[SUP] 4 [/SUP], Dennis Shung[SUP] 5 [/SUP], Alexander Tong[SUP] 2 [/SUP], Carolina Lucas[SUP] 3 [/SUP], Jon Klein[SUP] 3 [/SUP], Daniel B Burkhardt[SUP] 6 [/SUP], Scott Gigante[SUP] 7 [/SUP], Abhinav Godavarthi[SUP] 8 [/SUP], Bastian Rieck[SUP] 9 [/SUP], Benjamin Israelow[SUP] 3 10 [/SUP], Michael Simonov[SUP] 5 [/SUP], Tianyang Mao[SUP] 3 [/SUP], Ji Eun Oh[SUP] 3 [/SUP], Julio Silva[SUP] 3 [/SUP], Takehiro Takahashi[SUP] 3 [/SUP], Camila D Odio[SUP] 5 [/SUP], Arnau Casanovas-Massana[SUP] 11 [/SUP], John Fournier[SUP] 10 [/SUP], Yale IMPACT Team; Shelli Farhadian[SUP] 10 [/SUP], Charles S Dela Cruz[SUP] 12 13 [/SUP], Albert I Ko[SUP] 10 11 [/SUP], Matthew J Hirn[SUP] 14 15 [/SUP], F Perry Wilson[SUP] 16 [/SUP], Julie G Hussin[SUP] 4 17 [/SUP], Guy Wolf[SUP] 18 19 [/SUP], Akiko Iwasaki[SUP] 3 20 [/SUP], Smita Krishnaswamy[SUP] 21 22 [/SUP]
Collaborators, Affiliations
Abstract
As the biomedical community produces datasets that are increasingly complex and high dimensional, there is a need for more sophisticated computational tools to extract biological insights. We present Multiscale PHATE, a method that sweeps through all levels of data granularity to learn abstracted biological features directly predictive of disease outcome. Built on a coarse-graining process called diffusion condensation, Multiscale PHATE learns a data topology that can be analyzed at coarse resolutions for high-level summarizations of data and at fine resolutions for detailed representations of subsets. We apply Multiscale PHATE to a coronavirus disease 2019 (COVID-19) dataset with 54 million cells from 168 hospitalized patients and find that patients who die show CD16[SUP]hi[/SUP]CD66b[SUP]lo[/SUP] neutrophil and IFN-γ[SUP]+[/SUP] granzyme B[SUP]+[/SUP] Th17 cell responses. We also show that population groupings from Multiscale PHATE directly fed into a classifier predict disease outcome more accurately than naive featurizations of the data. Multiscale PHATE is broadly generalizable to different data types, including flow cytometry, single-cell RNA sequencing (scRNA-seq), single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq), and clinical variables.
. 2022 Feb 28.
doi: 10.1038/s41587-021-01186-x. Online ahead of print.
Multiscale PHATE identifies multimodal signatures of COVID-19
Manik Kuchroo[SUP] #[/SUP][SUP] 1 [/SUP], Jessie Huang[SUP] #[/SUP][SUP] 2 [/SUP], Patrick Wong[SUP] #[/SUP][SUP] 3 [/SUP], Jean-Christophe Grenier[SUP] 4 [/SUP], Dennis Shung[SUP] 5 [/SUP], Alexander Tong[SUP] 2 [/SUP], Carolina Lucas[SUP] 3 [/SUP], Jon Klein[SUP] 3 [/SUP], Daniel B Burkhardt[SUP] 6 [/SUP], Scott Gigante[SUP] 7 [/SUP], Abhinav Godavarthi[SUP] 8 [/SUP], Bastian Rieck[SUP] 9 [/SUP], Benjamin Israelow[SUP] 3 10 [/SUP], Michael Simonov[SUP] 5 [/SUP], Tianyang Mao[SUP] 3 [/SUP], Ji Eun Oh[SUP] 3 [/SUP], Julio Silva[SUP] 3 [/SUP], Takehiro Takahashi[SUP] 3 [/SUP], Camila D Odio[SUP] 5 [/SUP], Arnau Casanovas-Massana[SUP] 11 [/SUP], John Fournier[SUP] 10 [/SUP], Yale IMPACT Team; Shelli Farhadian[SUP] 10 [/SUP], Charles S Dela Cruz[SUP] 12 13 [/SUP], Albert I Ko[SUP] 10 11 [/SUP], Matthew J Hirn[SUP] 14 15 [/SUP], F Perry Wilson[SUP] 16 [/SUP], Julie G Hussin[SUP] 4 17 [/SUP], Guy Wolf[SUP] 18 19 [/SUP], Akiko Iwasaki[SUP] 3 20 [/SUP], Smita Krishnaswamy[SUP] 21 22 [/SUP]
Collaborators, Affiliations
- PMID: 35228707
- DOI: 10.1038/s41587-021-01186-x
Abstract
As the biomedical community produces datasets that are increasingly complex and high dimensional, there is a need for more sophisticated computational tools to extract biological insights. We present Multiscale PHATE, a method that sweeps through all levels of data granularity to learn abstracted biological features directly predictive of disease outcome. Built on a coarse-graining process called diffusion condensation, Multiscale PHATE learns a data topology that can be analyzed at coarse resolutions for high-level summarizations of data and at fine resolutions for detailed representations of subsets. We apply Multiscale PHATE to a coronavirus disease 2019 (COVID-19) dataset with 54 million cells from 168 hospitalized patients and find that patients who die show CD16[SUP]hi[/SUP]CD66b[SUP]lo[/SUP] neutrophil and IFN-γ[SUP]+[/SUP] granzyme B[SUP]+[/SUP] Th17 cell responses. We also show that population groupings from Multiscale PHATE directly fed into a classifier predict disease outcome more accurately than naive featurizations of the data. Multiscale PHATE is broadly generalizable to different data types, including flow cytometry, single-cell RNA sequencing (scRNA-seq), single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq), and clinical variables.