Pinned Tweet
Most tumours are never profiled with spatial transcriptomics, but nearly all are imaged. DeepSpot-M bridges that gap and as a first demonstration, we used it to build a virtual spatial transcriptomics atlas of 28,664 TCGA slides across 32 cancers, the largest resource of its kind
1
0
3
394
Kalin Nonchev retweeted
There is a lot of talk about what AI might do for medicine one day, and then there is Kalin, whose model already opens an expensive cancer research method to labs that could never afford the equipment.
One of cancer research’s fastest growing methods is up 13x since 2021, yet fewer than 1% of its papers come from lower-income countries due to high costs.

Kalin Nonchev’s (@nonchevk) DeepSpot tackles this gap by predicting gene maps from standard tissue slides available in labs worldwide.
0
1
7
342
Thanks to @crunchDAO for highlighting the transformative potential of AI in spatial biology and our recent perspective on the unequal geography of the field.
One of cancer research’s fastest growing methods is up 13x since 2021, yet fewer than 1% of its papers come from lower-income countries due to high costs.

Kalin Nonchev’s (@nonchevk) DeepSpot tackles this gap by predicting gene maps from standard tissue slides available in labs worldwide.
3
0
3
129
Replying to @nonchevk
✨Read our perspective: auroraomics.org/news/the-unequ…
🔬 Working in academic research? Try Aurora Direct and explore virtual spatial gene expression from your own histology images. auroraomics.org/log-in
0
0
0
3
Replying to @nonchevk
✨Read our perspective: auroraomics.org/news/the-unequ…
🚀 Access Aurora Direct: auroraomics.org/log-in
0
0
1
2
Kalin Nonchev retweeted
One of cancer research’s fastest growing methods is up 13x since 2021, yet fewer than 1% of its papers come from lower-income countries due to high costs.

Kalin Nonchev’s (@nonchevk) DeepSpot tackles this gap by predicting gene maps from standard tissue slides available in labs worldwide.
15
2
72
16,651
🌍 Virtual spatial transcriptomics goes global!

An exciting new milestone! Researchers from every inhabited continent have now analyzed virtual spatial transcriptomics data generated using our DeepSpot-M model.

auroraomics.org/news/virtual-s…
1
0
1
36
Replying to @nonchevk
🔬 Working in academic research? Try Aurora Direct and explore predicted spatial gene expression from your own histology images. auroraomics.org/log-in
1
0
0
11
To mark World Breast Cancer Day, we’ve decided to dedicate the additional compute capacity we’ve unlocked through recent improvements to Aurora to support breast cancer researchers with virtual spatial transcriptomics from H&E images. auroraomics.org/news/breast-ca…
1
0
1
77
Replying to @nonchevk
The breast cancer section used for the independent comparison is taken from 10x Genomics data at 10xgenomics.com/datasets/human…
1
0
0
11
Next Wednesday I’ll be at @BioTechX_ in Basel presenting our new models for virtual spatial transcriptomics from histology during the poster session from 13:25–14:25
If you’d like to learn more, stop by the poster or message me here or via the conference app to arrange a meeting
1
0
1
59
🌍 Less than 1% of spatial transcriptomics papers have a first or last author affiliated with a research organisation in a low-income economy. Compared to 17% in histopathology. auroraomics.org/news/the-unequ…
1
0
0
71
Replying to @nonchevk
We looked at the geography of spatial transcriptomics research, the scale of the gap, and why virtual spatial transcriptomics from histology could help broaden access and enable discoveries.
1
0
0
15