dalloliogm

Here is my approach to the CAFA6 challenge: predicting protein function from sequence. It is based on converting everything into numbers, using ESM2 for protein sequences, and an LLM for Gene Ontology terms. gmdbioinformatics.substack.com/p/my-strategy-… #CAFA6 #ESM2 #deepLearning #bioinformatics

The winners of the Arc Virtual Cell Challenge have been announced! Did a foundation model outperform regressions? Did the leaderboard hacking matter in the end? Here are my thoughts. gmdbioinformatics.substack.com/p/arc-virtual-… #ArcVirtualCellChallenge #VirtualCellChallenge

Proud to have contributed to this bioArXiv paper, presenting an Agentic System for Cell Type annotation in single-cell open.substack.com/pub/gmdbioinfo… #singlecell #bioinformatics #ngs #cytetype

My third article on the Arc Virtual Cell Challenge. This time, we learn how to train our very first Virtual Cell Model. gmdbioinformatics.substack.com/p/arc-virtual-… #ArcInstitute #VirtualCellChallenge #BiologicalModelling #FoundationModels #MachineLearning #ML

My second post on the Arc Virtual Cell Challenge. The challenge’s Discord forums are in turmoil. Some participants have discovered a trick to get to the top of the leaderboard. gmdbioinformatics.substack.com/p/arc-virtual-… #arc_virtual_cell_challenge #foundation_models

@wildtypehuman @RelationRx @BlackHC Thanks for sharing this. The blog post is very useful to understand the context.

@RelationRx @BlackHC And here's the blog post if you want to know what's AL all about: wildtypehuman.substack.com/p/making-ai-us…

Much of the active learning literature avoids modelling the bias and noise of the underlying experimental system...

With Paul Scherer of @RelationRx and Andreas Kirsch (@BlackHC) we give our take on how to exploit a very simple relationship (the bias-variance tradeoff) to build a class of general AL strategies. arxiv.org/abs/2509.04363

My 8yo son and I developed a maths game called Times Tables Animals, featuring cute bunnies, maths, and volcanoes. Here is our story. dalloliogm.github.io/Times-Table-An… @Kilo_Code @TTRockStars #llm #parentinghacks

Absolutely fascinating paper! Some of the best examples of learning from evolution comes from non-human GWAS. A missense mutation (discovered previously via a GWAS) in FREP1 encoding a mosquito gut protein provides near complete resistance to malarial parasite infection. The beauty of genetics is it not only pinpoints the protein that needs to be targeted, it also tells what function of the protein could be targeted without killing the entire protein (which will impose fitness costs) In this case, the FREP1 protein is critical for normal physiological functions of mosquito host. But the missense mutation selectively affects the function that is hijacked by the parasites to invade the mosquito gut. Here the authors propose driving this protective allele in to the wild mosquito population (instead of deleting the gene which is lethal and might not help the gene drive) using modern genetic tools would be an effective approach to eliminate malaria. Li, Dong, et al. Nature nature.com/articles/s4158…

Am I the slowest user on Kaggle? My post from the Kaggle Hackathon 2025. @dalloliogm/am-i-the-slowest-kaggle-user-ever-457c4a382803" target="_blank" rel="nofollow noopener">medium.com/@dalloliogm/am…

How we "guessed" the Pope using network science: inside the cardinal network. A study by me, Beppe Soda and Alessandro Iorio. Article: unibocconi.it/en/news/networ… @Unibocconi

@0xCharlota @realEstateTrent Is this what twitter is, nowadays?

@realEstateTrent Wild how many replies here are like “oh your wife should totally pause her career so she can take care of the dinner mess and meal prep”

My controversial take on the Medallion Architecture from Databricks. It is a framework for managing user expectations regarding data requests. @dalloliogm/managing-expectations-with-the-medallion-architecture-71be205fb98e" target="_blank" rel="nofollow noopener">medium.com/@dalloliogm/ma…

For those who haven't come across it yet, here's a handy trick to discuss an entire GitHub repo with an LLM: => Just replace "github" with "gitingest" in the url, and you get the whole repo as a single string that you can then paste in your LLMs

MLflow is the missing piece for ensuring reproducibility of analysis in Bioinformatics. However, not many people in the field are familiar with it. Here I provide an example based on Differential Expression. @dalloliogm/mlflow-for-bioinformatics-the-missing-piece-for-reproducibility-of-results-ad8764dbcc83" target="_blank" rel="nofollow noopener">medium.com/@dalloliogm/ml…… #Bioinformatics #mlflow

Foundation Models for Biology are going to revolutionize Bioinformatics in 2025. I wrote a short article to explain what they are and how they can be used. #llm #bioinformatics medium.com/p/84d74b12b7a7

@tangming2005 Do you use chatGPT or other LLMs to help you writing the posts? Thanks for all the great material

Pre-scheduling a long post a day is more time-consuming. I just spent over 3 hours finishing my coming week's posts. It is not easy. My wife is taking a bath for the kids and I am here drafting all the posts.

tidyplots Time to say goodbye to ggplot2?🫡 "a significant reduction of code complexity" vs ggplot2 cran.r-project.org/web/packages/t… @JanBroderEngler bioRxiv 2024 biorxiv.org/content/10.110…

@vsbuffalo Summarize the publications section. Something like "published X+ number of publications in top scientific journals. Made contributions to <field name> that helped understand <things>. Essentially, you to need to explain the impact of your work to people who are not in the field.

Is there a good guide to converting an academic CV to an industry CV? Anyone with a good CV for biotechs wanna share theirs? I find it entertaining that Claude suggested shortening my publication section significantly. Different priorities!

Winner: best software name