ChloeXWang

29 posts

ChloeXWang

ChloeXWang

@ChloeXWang1

CS PhD student at the WangLab @ U of T. Working on ML applications in single-cell research.

Katılım Temmuz 2023
44 Takip Edilen152 Takipçiler
Sabitlenmiş Tweet
ChloeXWang
ChloeXWang@ChloeXWang1·
1/7 First of all, big shoutout to co-authors on modeling (@MKarimzade, @neal_ravindra, @RexMa9, @HAOTIANCUI1, @LeeTaliq), huge appreciation to data generation (Lexi, @alerasool, Adam) and bioinformatics team (@_annhuang), and leadership for vision and direction (@BoWang87, @inCiChu)! Preprint is now live on bioRxiv: biorxiv.org/content/10.648… All models start from high-quality data.
Bo Wang@BoWang87

Our X-cell is up at @biorxiv_bioinfo ! Read our full paper at biorxiv.org/content/10.648… Part of the data and the model weights will be shared soon. stay tuned!

English
1
11
33
6.8K
ChloeXWang retweetledi
Bo Wang
Bo Wang@BoWang87·
It was super fun to chat with @abhinadduri and @genophoria at @arcinstitute about our BioReason series!
Bo Wang tweet media
Abhinav Adduri@abhinadduri

We @arcinstitute, @UHN, and @VectorInst recently released out BioReason-Pro, a multimodal reasoning LLM for protein function prediction, trained via SFT on synthetic reasoning traces and subsequent RL. I had a chance to interview @BoWang87 and @genophoria on their vision for the work and what comes next. Was fun to pick their brains on the bio! Check out the interview: youtube.com/watch?v=uZx3nU…

English
1
4
25
3.8K
ChloeXWang
ChloeXWang@ChloeXWang1·
6/7 We use test-time adaptation (TTA) to address distributional shift in zero-shot inference and align the model to new contexts: • adapt on control cells only • no perturbation data leakage
ChloeXWang tweet media
English
1
0
5
164
ChloeXWang
ChloeXWang@ChloeXWang1·
1/7 First of all, big shoutout to co-authors on modeling (@MKarimzade, @neal_ravindra, @RexMa9, @HAOTIANCUI1, @LeeTaliq), huge appreciation to data generation (Lexi, @alerasool, Adam) and bioinformatics team (@_annhuang), and leadership for vision and direction (@BoWang87, @inCiChu)! Preprint is now live on bioRxiv: biorxiv.org/content/10.648… All models start from high-quality data.
Bo Wang@BoWang87

Our X-cell is up at @biorxiv_bioinfo ! Read our full paper at biorxiv.org/content/10.648… Part of the data and the model weights will be shared soon. stay tuned!

English
1
11
33
6.8K
ChloeXWang retweetledi
Bo Wang
Bo Wang@BoWang87·
2026 may be the year AI starts to truly reason about biology. AlphaFold helped close the sequence → structure gap. The next frontier is sequence → functions. Today, together with @genophoria and the team at @arcinstitute , we’re releasing BioReason-Pro — the first multimodal reasoning model for protein function prediction.
Bo Wang tweet media
English
15
76
304
69.8K
Adib
Adib@adibvafa·
@ChloeXWang1 Impressive work, congratulations!
English
1
0
0
58
ChloeXWang retweetledi
Bo Wang
Bo Wang@BoWang87·
Tahoe-100M is a groundbreaking dataset for AI-driven drug discovery, setting new standards in both scale and quality. Its unprecedented cell type diversity, extensive drug coverage, large number of replicates, and rich metadata will drive the development of next-generation AI tools. High-quality datasets are rare yet crucial for enhancing the predictive power of foundation models—Tahoe-100M is exactly what single-cell foundation models have been waiting for. Congratulations @nalidoust and @vevo_ai !
Nima Alidoust@nalidoust

Historic day for builders in bio: We’ve open-sourced @vevo_ai’s #Tahoe100M, largest single-cell atlas ever—by a wide margin—as the inaugural contribution to @arcinstitute’s Virtual Cell Atlas, ready for download today. A leap forward for AI models of cells & drug discovery. 🧵

English
3
24
115
14.2K
ChloeXWang
ChloeXWang@ChloeXWang1·
12/12 We demonstrate improved performance in both tasks by enriching the input to the Tangram pipeline with fine-tuned scGPT embeddings, highlighting the versatility of scGPT-spatial in supporting both single-cell and Visium references.
ChloeXWang tweet media
English
1
0
3
124
ChloeXWang
ChloeXWang@ChloeXWang1·
11/12 We further explored the utility of scGPT-spatial embeddings in enhancing spatial transcriptomic data by deconvoluting Visium spots into cell type compositions and imputing missing genes in Xenium slides with Tangram pipeline.
ChloeXWang tweet media
English
1
0
2
188

Keşfet

@abhinadduri @genophoria @arcinstitute @MKarimzade @neal_ravindra @RexMa9 @HAOTIANCUI1 @LeeTaliq