From 30958ea47892dc5b50bc92b770ba724699d07fe1 Mon Sep 17 00:00:00 2001 From: Paul Gueguen <34238952+p-gueguen@users.noreply.github.com> Date: Thu, 10 Oct 2024 09:52:28 +0200 Subject: [PATCH] Update README.md --- README.md | 71 +++++++++++++++++++++++++++++-------------------------- 1 file changed, 38 insertions(+), 33 deletions(-) diff --git a/README.md b/README.md index 304fd4a..f5d0a24 100644 --- a/README.md +++ b/README.md @@ -1,37 +1,25 @@ -# Spatial Transcriptomics tools +# Spatial analyses > **Smaller** | Cell types → Cell modules - cell neighborhoods → Niches/tissue domain | **Larger** +> # Spatial analyses -- [Benchmark](https://www.biorxiv.org/content/10.1101/2023.03.22.533802v3.full.pdf) claims that [RCTD](https://github.com/dmcable/spacexr) and [Cell2location](https://github.com/BayraktarLab/cell2location) are the best -- [Benchmark](https://www.nature.com/articles/s41592-024-02215-8) on spatial clustering -- Other review [here](https://www.nature.com/articles/s41467-023-37168-7#Fig2), Cell2location again -- Cell2location is the best in the [open problems](https://openproblems.bio/results/spatial_decomposition/) +**Other ressource** → https://github.com/crazyhottommy/awesome_spatial_omics -[A standard for sharing spatial transcriptomics data](https://www.cell.com/cell-genomics/fulltext/S2666-979X(23)00171-4) +## General advice -- [Detection of SVGs](https://www.nature.com/articles/s41467-024-44835-w) with PROST -- [Super resolution 10X Visium with TESLA](https://www.cell.com/cell-systems/pdf/S2405-4712(23)00084-4.pdf) -- Super resolution Visium with istar https://www.nature.com/articles/s41587-023-02019-9 -- [Subspot resolution with BayesSPACE](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8763026/) -- https://github.com/crazyhottommy/awesome_spatial_omics +- This paper recommends [cell volume normalization](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-024-03303-w#Fig7) for imaging-based techniques, especially when the list of probes is small ## General Tools -- Best practices - -[Best Practices for Spatial Transcriptomics Analysis with Bioconductor - 2  Spatial transcriptomics](https://lmweber.org/BestPracticesST/chapters/spatial-transcriptomics.html) - -- **Doing gene imputation is [not recommended](https://github.com/BayraktarLab/cell2location/issues/379)** +- [Best practices Bioconductor](https://lmweber.org/PrinciplesSTA/devel/) - scverse / [squidpy](https://squidpy.readthedocs.io/en/stable/) - [Giotto](https://giottosuite.readthedocs.io/en/latest/) suite - [Vitessce](http://vitessce.io/) - [Voyager](https://github.com/pachterlab/voyager) from Pachter lab -- [scran](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-024-03241-7) for normalization - Multiple sample analysis with [BASS](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-024-03361-0) -- SpaVAE to do everything, reduction, visualization, clustering, batch integration, denoising, differential expression, spatial interpolation, and resolution enhancement. -- This paper recommends [cell volume normalization](https://genomebiology.biomedcentral.com/articles/10.1186/s13059-024-03303-w#Fig7). +- [SpaVAE](https://github.com/ttgump/spaVAE) to do everything: reduction, visualization, clustering, batch integration, denoising, differential expression, spatial interpolation, and resolution enhancement. ### QC @@ -44,18 +32,22 @@ - Able to do supersampling / restoration with [cellpose 3](https://www.biorxiv.org/content/10.1101/2024.02.10.579780v2) - [DeepCell](https://github.com/vanvalenlab/deepcell-tf?tab=readme-ov-file) - [Best than SOTA segmentation - Nature Methods 2024 - Bo Wang](https://www.nature.com/articles/s41592-024-02233-6#Fig3) +- [Bin2Cell](https://www.biorxiv.org/content/10.1101/2024.06.19.599766v1) to segment VisiumHD data - [Proseg](https://www.biorxiv.org/content/10.1101/2024.04.25.591218v1.full.pdf) -- [Bin2Cell](https://www.biorxiv.org/content/10.1101/2024.06.19.599766v1) to segment Visium data -- https://bioimage.io/#/ segmentation repo - [ComSeg](https://www.nature.com/articles/s42003-024-06480-3) uses transcripts to segment - [FICTURE](https://www.nature.com/articles/s41592-024-02415-2) +- https://bioimage.io/#/ segmentation repo + +### SVGs + +- [Detection of SVGs](https://www.nature.com/articles/s41467-024-44835-w) with PROST ### Cell niches - [BANKSY](https://www.nature.com/articles/s41588-024-01664-3) **unifies cell typing and tissue domain segmentation for scalable spatial omics data analysis** -- https://twitter.com/Cancer_dynamics/status/1770811468578971905?t=sZhz7UBO-VE7cqK97rP8sA&s=19 staryfish -- https://buff.ly/49Oc0M2 TISSUE -- Hierarchy of Cell niches with CellCharter. Looking at the most stable solutions. There is a hierarchy of niches at different layers/hierarches. +- [staryfish](https://twitter.com/Cancer_dynamics/status/1770811468578971905?t=sZhz7UBO-VE7cqK97rP8sA&s=19) +- [TISSUE](https://buff.ly/49Oc0M2) +- [CellCharter](https://www.nature.com/articles/s41588-023-01588-4). Looking at the most stable solutions. There is a hierarchy of niches at different layers/hierarches. - [SpatialGLUE](https://www.nature.com/articles/s41592-024-02316-4) for cell niches **with multi-omics** - [smoothclust](https://github.com/lmweber/smoothclust) from lukas weber - [SpaTopic](https://www.science.org/doi/10.1126/sciadv.adp4942) @@ -65,8 +57,7 @@ - [Sprawl](https://elifesciences.org/reviewed-preprints/87517) - [Bento](https://www.notion.so/Spatial-analyses-0d451532f4c64fc599cb6ceb469ab523?pvs=21) - [FISHfactor](https://academic.oup.com/bioinformatics/article/39/5/btad183/7114027) -- https://www.nature.com/articles/s41467-024-49457-w -- [FICTURE](https://www.nature.com/articles/s41592-024-02415-2) +- [**InSTAnT**](https://www.nature.com/articles/s41467-024-49457-w) ### CCC @@ -79,7 +70,7 @@ ### Spatial trajectories -- spaTrack +- [spaTrack](https://www.biorxiv.org/content/10.1101/2023.09.04.556175v2) ### TF/GRN @@ -89,15 +80,29 @@ - PASTE/[PASTE2](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9881963/) -### Foundation models +### Super resolution -- https://huggingface.co/owkin/phikon-v2 -- Bioptimus: https://huggingface.co/bioptimus/H-optimus-0 +- [Super resolution 10X Visium with TESLA](https://www.cell.com/cell-systems/pdf/S2405-4712(23)00084-4.pdf) +- [Super resolution Visium with istar](https://www.notion.so/Spatial-analyses-0d451532f4c64fc599cb6ceb469ab523?pvs=21) +- [Subspot resolution with BayesSPACE](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8763026/) -# Datasets +### Denoising -- Interesting datasets embedding CNN + human features from DeepCell https://exploredata.deepcell.com/cell-visualizations/9/versions/1 +- **Doing gene imputation is [not recommended](https://github.com/BayraktarLab/cell2location/issues/379)** +- [Sprod](https://www.nature.com/articles/s41592-022-01560-w#Fig2) -Benchmark +## Benchmarks - https://www.biorxiv.org/content/10.1101/2024.04.03.586404v1.full +- [Benchmark](https://www.biorxiv.org/content/10.1101/2023.03.22.533802v3.full.pdf) claims that [RCTD](https://github.com/dmcable/spacexr) and [Cell2location](https://github.com/BayraktarLab/cell2location) are the best +- [Benchmark](https://www.nature.com/articles/s41592-024-02215-8) on spatial clustering +- Other review [here](https://www.nature.com/articles/s41467-023-37168-7#Fig2), Cell2location again +- Cell2location is the best in the [open problems](https://openproblems.bio/results/spatial_decomposition/) + +# Datasets + +### Foundation models + +- https://huggingface.co/owkin/phikon-v2 +- Bioptimus: https://huggingface.co/bioptimus/H-optimus-0 +- Interesting datasets embedding CNN + human features from DeepCell https://exploredata.deepcell.com/cell-visualizations/9/versions/1