Sign in

Jieran Sun

@onesinglecell.bsky.social
387 followers 134 following 7 posts

I treat every #singleCell gently. | PhD in Comp Bio at Lausanne University Hospital. #OpenScience #SpatialTranscriptomics

PostsRepliesMedia
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
This work is a result of SpaceHack2.0. Shout out to my amazing collaborators, especially Kirti Biharie, Peiying Cai, and Niklas Müller-Bötticher, and to the great support from @rgottardo.bsky.social, Brian Long, Ahmed Mahfouz, @markrobinsonca.bsky.social, and @naveed-ishaque.bsky.social(7/7)
000
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
Takeaway: Instead of a single clustering rep, try consensus + entropy, and get experts involved early. Everything is open-source. (6/7) 📄 doi.org/10.1038/s415... 💻 github.com/SpatialHacka...
doi.org
Beyond benchmarking: an expert-guided consensus approach to spatially aware clustering - Nature Methods
Benchmarking spatially aware clustering is challenging. SACCELERATOR is an open-source, extensible framework that formalizes expert-guided spatial clustering analysis.
110
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
We tested this with tissue experts. In the mouse thalamus, it helped correct the VPM cluster boundary. In colon cancer tissue, it helped two pathologists find a tumor boundary that wasn't in any existing reference. (5/7)
100
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
Our fix: stop ranking against shaky ground truth. Instead, we build a consensus from methods in the pipeline, and highlight where they disagree (cross-method entropy). Disagreement often marks real biological transitions or fuzzy boundaries. (4/7)
100
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
A bigger problem is the "ground truth" in benchmarks: they are manual histology annotations or computational clusters, used as the universal answer for every method. But these labels are subjective, biased toward one resolution, and miss multi-scale biological features. (3/7)
100
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
When benchmarking spatially-aware clustering methods, we found that their performance is reported inconsistently across papers, heavily affected by parameters and technologies, and most methods don't scale well to larger datasets. (2/7)
100
Jieran Sun @onesinglecell.bsky.social · 13/09/2026
Our paper is out in Nature Methods! 🧑‍💻 There are 50+ methods for spatially-aware clustering on spatial transcriptomics data. How should we use them? A thread 🧵 (1/7) 📄 doi.org/10.1038/s415... #spatialtranscriptomics #bioinformatics #computationalbiology
doi.org
Beyond benchmarking: an expert-guided consensus approach to spatially aware clustering - Nature Methods
Benchmarking spatially aware clustering is challenging. SACCELERATOR is an open-source, extensible framework that formalizes expert-guided spatial clustering analysis.
1124
Reposted by Jieran Sun
Siyuan Luo @siyuanluo.bsky.social · 05/12/2024
Excited to share our preprint on selecting validation metrics for single-cell and spatial omics! Explore how to evaluate embeddings, graphs, clustering, and spatial domains with biological relevance in mind. Plus, discover poem, our R package with new spatially-aware metrics! doi.org/10.1101/2024...
doi.org
On metrics for subpopulation detection in single-cell and spatial omics data
Benchmarks are crucial to understanding the strengths and weaknesses of the growing number of tools for single-cell and spatial omics analysis. A key task is to distinguish subpopulations within compl...
153
Reposted by Jieran Sun
Leander @le-and-er.bsky.social · 21/11/2024
Super excited to share our Human Neural Organoid Atlas, now out in Nature! Led by @zhisonghe.bsky.social @josch1.bsky.social, and myself, this resource was created from 36 scRNA-seq datasets—totalling over 1.7 million cells! 🔬✨
 www.nature.com/articles/s41...
 Find out how it can serve you ⏬
 🧵1/8
nature.com
An integrated transcriptomic cell atlas of human neural organoids - Nature
A human neural organoid cell atlas integrating 36 single-cell transcriptomic datasets shows cell types and states and estimates transcriptomic similarity between primary and organoid counterparts, sho...
37723
Reposted by Jieran Sun
Mike Schatz @mikeschatz.bsky.social · 25/11/2024
Sometimes you just have to persevere through negative reviews. Reviewers, please be kind :) #methodsmatter bit.ly/gs2-scholar
1425541