scRNAseq cell type annotation is notoriously messy. Despite so many algorithms, most researchers still rely on manual annotations using marker genes

In a new preprint accepted at ICML GenAI Bio Workshop, we ask if reasoning LLMs (DeepSeek-R1) can help with cell type annotation🧵 Driven by @samwang36 & @RunziTan97745 & with @BoWang87, we benchmarked DeepSeek-R1-0528 on zero-shot scRNAseq cell type annotation against non-reasoning LLMs, classifiers & foundation models . What we found surprised us👇biorxiv.org/content/10.110…

Impressive advancement in Computational Pathology.

A new multimodal foundation model by @AI4Pathology trained on 47,000 paired histology & genomics, which beautifully shows the multi-modal power of images & DNA & RNA

Even though patient genomic data is rare, it's so powerful 🧵 First, why is this model so important?

To my view, THREADS is the closest we have today to a cancer-level patient-centric foundation model.

It beautifully integrates lots of images, DNA & RNA - 3 data modalities providing critical orthogonal information about cancerous tissues

Many people wonder what is the scientific evidence behind what @sama & Larry Ellison said today
at The White House: that AI will cure cancer.

Truth is that this is not a hype. The potential of AI to accelerate cancer discoveries like never before is enormous.

Here’s why🧵 To start with: cancer is a very difficult problem. Funded with several billion dollars from the US government alone over the past few years, cancer survival has only marginally improved & incidence is actually increasing in younger people. That’s not good, in fact it’s really bad

https://twitter.com/simocristea/status/1879923278770241838

Cancer statistics in 2025 🇺🇸- new report 🚨

Cancer is becoming a new disease.

We're seeing a fundamental shift in who gets cancer, moving from a predominantly male, elderly disease to one that increasingly affects women and younger people.

Key highlights 🧵
1. New Cancer Cases and Deaths in 2025:

An estimated 2,041,910 new cancer cases will be diagnosed.

Approximately 618,120 cancer deaths will occur.

This equals about 5,600 new cases and 1,700 deaths per day.

To end 2023, I’ll share one of the most insightful & well-written papers I read in 2023.

This study @Nature links *spatial* tumor organization to immunotherapy response in breast cancer.

Immunotherapy is our strongest weapon against cancer. We need to understand it better.
🧵🧵 Long thread ahead, going deep into the molecular workings of breast cancer immunotherapy.

TL;DR:
1. Cancer–immune interactions & proliferative fractions predict immunotherapy response
2. Both pre-treatment & on-treatment predictors
3. Immunotherapy remodels the microenvironment

New: the monthly roller-coaster through October’s coolest life science papers is here 🚀🧬

3-sentence summaries of papers on evolution, single cell methodologies, genetic screens & more.

And, only for October, an educational video on fighting cancer🤺 as a bonus.

Enjoy 3x10! 1. Assembly theory (Sharma et al., Nature)

The most (in)famous paper I read this month proposes a new framework (assembly theory, that is) to explain basically everything… or, more specifically, “to unify descriptions of evolutionary selection across physics and biology” 1/3

The human genome is gradually unravelling its secrets 🎁

AlphaMissense model @ScienceMagazine: one more path lit up by deep learning in exploring the code of life 🧬

We now know with high confidence if 89% of ALL missense variants are benign or pathogenic

Key contributions🧵🧵 First things first:

Missense variants = genetic variants (i.e DNA bases) that change the amino acid sequence (i.e groups of 3 bases, building blocks of proteins) in proteins.

Missense variants are more important than non-missense ones, as more likely to have functional impact.

3 amazing papers just out @Nature, the kind worth giving up sleep for🦉

Spatial multi-omics human maps:
-placenta: MIBI & DSP
-intestine: CODEX & snRNAseq & snATACseq
-kidney: Visium & scRNA & scATAC

After sequencing single cells, we are now finally putting them back together🧵 1. Placenta 1/5

Understanding the mysterious maternal processes that sustain embryo development is fascinating.

Mapping those spatially with proteins & mRNA to describe the maternal-fetal interface in the first half of pregnancy is really mind-blowing.

https://t.co/j14fWp8dZznature.com/articles/s4158…

Twitter messed up my previous thread, but this is too important to let it slide:

Here are (again) my summary & thoughts on early detection & an amazing work

Deep learning model trained on 9 million patient records in Denmark & US finds people at risk for pancreatic cancer
🧵🧵 In this thread, we'll discuss:

1. Context & significance of study
2. Datasets
3. Deep learning model
4. Model performance
5. Feature interpretability
6. Thoughts

And here's the link to the paper:

nature.com/articles/s4159…

Cancer is a terrible disease, and also one that we all know too well.

It is not a new problem, rather one that exists since thousands of years & is studied in unimaginable detail.

Then why do people still die of cancer?

Let's start understanding this by taking a step back. It’s 1938, and Public Health Services are advising people that detecting and treating cancers early will save their lives.

New preprint on scGPT: first foundation large language model for single cell biology, pretrained on 10 million cells.

Just as text is made of words, cells are characterized by genes

Some thoughts on how cool this is & why it challenges the status quo of single cell analysis🧵🧵 The self-attention transformer (chatGPT) is so successful in natural language processing (NLP)

But maybe single cell biology is not that different from NLP, as genes & cells correspond to words & sentences?

That's the framing of @BoWang87's awesome paper
biorxiv.org/content/10.110…

Only a matter of time before a paper formalized this exercise:

Automated #scRNAseq cell type annotation with GPT4, evaluated across five datasets, 100s of tissues & cell types, human and mouse.

A🧵below with my thoughts on how such tools will change how #Bioinformatics is done. I'll start with a quick summary of the paper, such that we're all on the same page.

(The paper is also a super quick read, literally only 3 pages of text, among which 1 is GPT prompts).

Here's the link to the preprint

biorxiv.org/content/10.110…

The paper I am sharing today is a thoughtful philosophical perspective from @sdomcke & @JShendure proposing a new organizational framework for single cell data, as an alternative to e.g Human Cell Atlas

Compelling read for both lovers❤️ & skeptics🤔 of single cell genomics

🧵🧵 This thread is organized as follows:

1️⃣ The need to organize Biology
2️⃣ How to organize cell types?
3️⃣ A consensus ontology
4️⃣ Structure & representation of the cell reference tree
5️⃣ Resolution of tree labels
6️⃣ Example tree
7️⃣ Human tree
8️⃣ Thoughts

Do you need to analyze Spatial Transcriptomics data, but are lost in the endless sea of methods?

Here's an explainer of the new @NatureComms paper benchmarking 18 spatial cellular deconvolution methods🧵🧵

nature.com/articles/s4146… This thread is organized as follows:

1️⃣ Intro to Spatial Transcriptomics
2️⃣ Intro to Cellular Deconvolution
3️⃣ Methods benchmarked
4️⃣ Datasets used (real & simulated)
5️⃣ Performance assessment
6️⃣ Benchmarking results
7️⃣ Accuracy
8️⃣ Robustness
9️⃣ Usability
🔟 Guidelines

🚨Our new study is out @CellReports!

We use single cell protein quantification & single cell FISH to map #spatial interactions in genetic mosaicism & tumor microenvironment in #Glioblastoma!

Wonderful collaboration w/ @janiszewska_lab @DalitEngelhardt
@Kacper_W_PhD

Deep dive👇 First, some context.

Glioblastoma (GBM) is one of the deadliest, most aggressive cancers that exist, with a median survival of only 15 months.

In GBM, 'single cell heterogeneity' are not simply buzzwords.

Rather, this immense heterogeneity is a main reason of treatment failure

I need to raise awareness about an important point in #scRNAseq data analysis, which, in my opinion, is not acknowledged enough:

‼️In practice, most cell type assignment methods will fail on totally novel cell types. Biological/expert curation is necessary!

Here's one example👇 Last year, together with @LabPolyak @harvardmed, we published a study in which we did something totally awesome: we experimentally showed how a TGFBR1 inhibitor drug 💊 prevents breast tumor initiation in two different rat models!

Here's a detailed thread on this paper:

https://twitter.com/simocristea/status/1600578512578035733

🚨New #SpatialTranscriptomics #Bioinformatics data resource out in @naturemethods.

SODB, a platform with >2,400 manually curated spatial experiments from >25 spatial omics technologies & interactive analytical modules.

This🧵will walk you through all the features of SODB [1/33] First, some background.

Spatial technologies complement classical genomics by also providing information about spatial context & tissue organization in:

- embriogenesis
- disease development
- normal tissue homeostasis

The field has exploded 🔥 in the past 2 years. [2/33]

Interested in how classical rule-based modular biology & #deeplearning fit together for the design of artificial proteins?

A new preprint combines these two modeling strategies to generate novel proteins!

Let's take a closer look at this innovative framework🧵👇 This method comes from the @MetaAI FAIR protein folks: @BrianHie, @salcandido, @ebetica, @OriKabeli, @proteinrosh, @nikismetanin, @TomSercu, @alexrives and is available as a preprint.

biorxiv.org/content/10.110…

🚨New milestone for #DeepLearning & life sciences in @NatureBiotech

Generating brand new functional proteins from scratch with large language models (e.g. #chatGPT)

Let’s understand this Transformers model used for protein design, how well it works & why this is important🧵👇 The very nice paper discussed in this thread comes from a team led by @nikhil_ai at Salesforce @SFResearch 👏

It was available online as a preprint since 2021.

biorxiv.org/content/10.110…

nature.com/articles/s4158…

Inspiring Symposium on Cancer Prevention @EACRnews

95% of cancer drugs fail. 94% do not improve life quality.

An ounce of prevention is worth a pound of cure. (B. Franklin)

Cancer prevention is tremendously difficult. But it is also necessary.

We need to shift our focus. How to move from developing cancer treatments to cancer prevention? @cohen_cyrille

How to change the single gene/ single mutation paradigm for holistic approaches considering multi-omics, lifestyle, exposure and cells as a whole? @AzraRazaMD

Graph Neural Networks (#GNNs) & their applications to life sciences are an exciting #DeepLearning area to discover!

But, to develop or apply GNN methods, we first need to understand the maths behind.

So, back to basics!

Here's a plain language summary of what's behind GNNs👇 This summary is based on @PetarV_93’s recent paper with introductory theoretical notions on Graph Neural Networks.

This resource is very much an introductory one.

arxiv.org/abs/2301.08210