Sign in

Ian Driver

@biodriver.bsky.social
235 followers 754 following 39 posts

Head of Bioinformatics at Gordian Biotechnology. Open science proponent. Runner. 🧬🖥️🏃🏻

PostsRepliesMedia
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 11/10/2026
Perturb-Seqr: Signature Search Engine Integrating Connectivity Mapping Resources www.biorxiv.org/content/10.64898/20…
011
Ian Driver @biodriver.bsky.social · 11/09/2026
Yeah the science account keeps working, but everything else opus 5 now refuses even basic science. The turned it on yesterday morning. The fact that is arbitrary and without comment or any communication makes it pretty much impossible to use Claude anymore.
011
Ian Driver @biodriver.bsky.social · 05/08/2026
Absolutely, there have been many use cases for me where i am the one who wrote the code to test against and or generated the data. In my area of expertise it remains useful. AI is a great help at scaling (workflow) and optimization, etc. I can just see my area of expertise rapidly changing.
010
Ian Driver @biodriver.bsky.social · 05/08/2026
As someone, like most scientists, who loves coming up with ideas and testing them, solving problems, coming up with hypothesis. Doing all of those thing faster is a lot of fun. For now I am solving the problems I find useful and interesting and as this paper shows, AI can help me do that.
010
Ian Driver @biodriver.bsky.social · 05/08/2026
Maybe the future is that what we view as code becomes like assembly and we trust the compiler-- now AI agents-- to handle it (as others have written about). As someone who likes writing and art and a deep understanding of problems that only comes with tackling them, I am uneasy.
140
Ian Driver @biodriver.bsky.social · 05/08/2026
goes unmaintained is huge. The ability to bring them back to life or make them useful I think is a real boon for science and open science. But the negatives are also clear: Who is writing code when ~no one is writing code? and Who is maintaining/reviewing code? See bsky.app/profile/jyn....
110
Ian Driver @biodriver.bsky.social · 05/08/2026
In this case, and many other scientific coding exercises, the best use case are versions of update this package, use this existing code to do something else, or optimize this package, etc. In scientific computing this is no small thing- the number of excellent packages/work by PhD students that
110
Ian Driver @biodriver.bsky.social · 05/08/2026
Wading into AI coding because I want to be able to run lots of single cell data on my laptop. We created Rustar (STAR solo) port because we had an exact package to test against. Fun and useful but begs the question of where AI takes coding when it doesn't have already existing code to test against?
252
Reposted by Ian Driver
bioRxiv Genomics @biorxiv-genomic.bsky.social · 10/07/2026
Isoform-level resolution in single-cell CRISPR screens reveals hidden functional consequences of gene perturbation www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 25/06/2026
fastQpick: scalable bootstrap and subsampling of FASTQ reads www.biorxiv.org/content/10.64898/20…
032
Ian Driver @biodriver.bsky.social · 16/06/2026
I got a new macbook and was curious if rapids style single-cell GPU works. The catch: PyTorch-MPS and MLX have no GPU SVD/eigendecomposition — so PCA & neighbors can't move to the GPU at all. The easy steps win; the steps that matter can't. github.com/iandriver/ap... #singlecell #bioinformatics
github.com
GitHub - iandriver/apple-gpu-singlecell-benchmark: Feasibility benchmark: can single-cell RNA-seq preprocessing (rapids-singlecell style) be GPU-accelerated on Apple Silicon/Metal? Measured on an M5 P...
Feasibility benchmark: can single-cell RNA-seq preprocessing (rapids-singlecell style) be GPU-accelerated on Apple Silicon/Metal? Measured on an M5 Pro. Spoiler: the Apple-GPU linalg gap blocks it....
020
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 16/06/2026
Expanding gene regulatory networks from transcriptome data through graphical modeling with heterogeneous priors www.biorxiv.org/content/10.64898/20…
031
Reposted by Ian Driver
Nature Methods @natmethods.nature.com · 15/06/2026
An Analysis evaluated the effect of pretraining dataset size and diversity on the performance of single-cell foundation models. @alandenadel.bsky.social @nfusi.bsky.social @lcrawford.bsky.social @avapamini.bsky.social www.nature.com/articles/s41...
nature.com
Evaluating the role of pretraining dataset size and diversity on single-cell foundation model performance - Nature Methods
The performance of single-cell foundation models is dependent on many factors. This study assesses the effect of the pretraining dataset’s size and diversity, revealing potential challenges in pursuin...
161
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 11/06/2026
HoloCell: A Generative Foundation Model for Holistic Cellular Modeling www.biorxiv.org/content/10.64898/20…
022
Reposted by Ian Driver
bioRxiv Genomics @biorxiv-genomic.bsky.social · 30/05/2026
Early Pulmonary Fibrosis is Defined by Niche- and Cell-Specific Molecular Programs www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
Rafael Irizarry @rafalab.bsky.social · 18/05/2026
A flaw in Person-Δ may be overstating progress in single-cell perturbation prediction models. Pearson warned about this in the 19th century: reusing the same controls induces spurious correlation. Split the controls, and much of the claimed prediction power fades: www.biorxiv.org/content/10.6...
biorxiv.org
0146
Reposted by Ian Driver
bioRxiv SysBio @biorxiv-sysbio.bsky.social · 14/05/2026
MechAInistic: An LLM-guided Multi-Agent System for Reasoning over Genome-Scale Constraint-Based Metabolic Models www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
bioRxiv Genomics @biorxiv-genomic.bsky.social · 08/05/2026
Spatial imprints of emergent cardiomyocyte states in the pressure-overloaded heart www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 06/05/2026
Building computational benchmarks: an Omnibenchmark reimplementation of a single-cell preprocessing pipeline evaluation www.biorxiv.org/content/10.64898/20…
031
Reposted by Ian Driver
Alejandro Montenegro @aemonten.bsky.social · 27/04/2026
"In this study, we present a genome-scale map of genetic interactions in the human haploid cell line HAP1, based on CRISPR-based perturbation of ∼4 million gene pairs. The resulting network comprises ∼89,000 high-confidence gene-gene interactions" doi.org/10.1016/j.ce...
0299
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 25/04/2026
Are Current AI Virtual Cell Models Useful for Scientific Discovery? www.biorxiv.org/content/10.64898/20…
031
Reposted by Ian Driver
#SingleCell preprints @prepub-singlecell.bsky.social · 18/04/2026
cellNexus: Quality control, annotation, aggregation and analytical layers for the Human Cell Atlas data #SingleCell 🧪🧬🖥️ www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
bioRxivpreprint @biorxivpreprint.bsky.social · 16/04/2026
Single-Cell and Tissue-Specific CRISPR Editing Analyses Unveil New Insights to Off-Targets and Translocations www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
Dominic Grün @dominicgrun.bsky.social · 14/04/2026
Our study on rare immune cell networks governing liver cholangitis is now online! Using multiplexed single-cell and spatial omics, we discovered a triad of cholangiocytes, dendritic cells and gdT cells regulating the inflammatory and fibrotic tissue response. Rare immune cells matter! rdcu.be/fc60M
rdcu.be
An immunobiliary single-cell atlas resolves crosstalk between type 2 conventional dendritic cells and γδ T cells in cholangitis
Nature Communications - Primary sclerosing cholangitis (PSC) is an autoimmune condition affecting the liver. Here, the authors generate an immunobiliary transcriptomic atlas and show that dendritic...
0125
Reposted by Ian Driver
Luciano Martelotto @lgmartelotto.bsky.social · 11/04/2026
We GoT a Filling…uuuuhoooo 𝗧𝗮𝗿𝗴𝗲𝘁𝗲𝗱 𝘀𝗲𝗾𝘂𝗲𝗻𝗰𝗶𝗻𝗴 𝗼𝗳 𝗺𝘂𝘁𝗮𝘁𝗶𝗼𝗻𝘀 𝘃𝗶𝗮 𝗥𝗡𝗔-𝘁𝗲𝗺𝗽𝗹𝗮𝘁𝗲𝗱 𝗴𝗮𝗽 𝗳𝗶𝗹𝗹𝗶𝗻𝗴 𝗼𝗳 𝗼𝗹𝗶𝗴𝗼𝗻𝘂𝗰𝗹𝗲𝗼𝘁𝗶𝗱𝗲𝘀 𝗳𝗼𝗿 𝘀𝗶𝗻𝗴𝗹𝗲-𝗰𝗲𝗹𝗹 𝗥𝗡𝗔-𝘀𝗲𝗾 www.biorxiv.org/content/10.6...
biorxiv.org
051
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 19/03/2026
Super Bloom: Fast and precise filter for streaming k-mer queries www.biorxiv.org/content/10.64898/20…
02213
Ian Driver @biodriver.bsky.social · 18/03/2026
PerturbAI out of stealth paper:
000
Reposted by Ian Driver
bioRxiv Genomics @biorxiv-genomic.bsky.social · 18/03/2026
Genome-scale functional mapping of the mammalian whole brain with in vivo Perturb-seq www.biorxiv.org/content/10.64898/20…
031
Reposted by Ian Driver
Nature Methods @natmethods.nature.com · 17/03/2026
CellVoyager: an AI agent that explores new biological hypotheses by autonomously analyzing scRNA-seq datasets. www.nature.com/articles/s41...
nature.com
CellVoyager: AI CompBio agent generates new insights by autonomously analyzing biological data - Nature Methods
CellVoyager is an artificial intelligence agent capable of exploring new biological hypotheses by autonomously analyzing single-cell RNA sequencing datasets and accounting for background information a...
161
Ian Driver @biodriver.bsky.social · 08/01/2026
Just the beginning of what we have been building and working towards at Gordian: www.businesswire.com/news/home/20...
businesswire.com
000
Reposted by Ian Driver
#SingleCell preprints @prepub-singlecell.bsky.social · 19/12/2025
DINOcell: Learning Generalizable Perturbation Effects through Self-Distillation #SingleCell 🧪🧬🖥️ www.biorxiv.org/content/10.64898/20…
021
Reposted by Ian Driver
bioRxivpreprint @biorxivpreprint.bsky.social · 12/12/2025
Reconstructing developmental and disease progression with sample-level embeddings www.biorxiv.org/content/10.64898/20…
011
Reposted by Ian Driver
Hannes Stark @hannes-stark.bsky.social · 26/10/2025
Excited to release BoltzGen which brings SOTA folding performance to binder design! The best part of this project is collaborating with a broad network of leading wetlabs that test BoltzGen at an unprecedented scale, showing success on many novel targets and pushing the model to its limits!
310341
Ian Driver @biodriver.bsky.social · 26/10/2025
We have thought about it, I did a lot of lung organoid work in my postdoc, just not convinced, for the disease we work on, that you can measure properties relevant enough to clinical outcomes in organoids
110
Ian Driver @biodriver.bsky.social · 26/10/2025
Pretty much what we do at Gordian Biotechnology, except we focus mostly on in vivo animal models. www.gordian.bio
gordian.bio
Gordian Biotechnology
Discovering new therapies for complex age-related disease.
110
Reposted by Ian Driver
bioRxiv SysBio @biorxiv-sysbio.bsky.social · 24/10/2025
Tahoe-x1: Scaling Perturbation-Trained Single-CellFoundation Models to 3 Billion Parameters www.biorxiv.org/content/10.1101/202…
011
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 16/10/2025
Perturbation-aware representation learning for in vivo genetic screens www.biorxiv.org/content/10.1101/202…
021
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 02/10/2025
FUSED: CROSS-DOMAIN INTEGRATION OF FOUNDATION MODELS FOR CANCER DRUG RESPONSE PREDICTION www.biorxiv.org/content/10.1101/202…
021
Reposted by Ian Driver
Darrell Kotton @kottond.bsky.social · 30/09/2025
Announcing our latest publication, just out today @NatureComms, advancing our understanding of interstitial lung disease. Congrats to co-senior authors Drs. Jeremy Katzen and Kostas Alysandratos and their teams @Penn @CReM_Boston. Check it out here: www.nature.com/articles/s41...
nature.com
Aberrant intermediate alveolar epithelial cells promote pathogenic activation of lung fibroblasts in preclinical fibrosis models - Nature Communications
The direct contribution of aberrant epithelial cells to lung fibrosis is largely unknown. The authors use murine and human systems to identify how these cells activate fibroblasts, and how reciprocal signals cause them to enter a profibrotic state.
074
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 25/09/2025
Domain-Invariant Feature Learning for Patient-Level Phenotype Prediction from Single-Cell Data www.biorxiv.org/content/10.1101/202…
011
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 24/09/2025
Single Cell Foundation Models Evaluation (scFME) for In-Silico Perturbation www.biorxiv.org/content/10.1101/202…
031
Reposted by Ian Driver
Nikolai Slavov @slavov-n.bsky.social · 22/09/2025
Some proteins are primarily regulated by one mechanism: RNA abundance, translation, or clearance. The regulation of most proteins is dominated by different regulatory mechanisms across cell types. Gratifyingly, this complex regulation defines simple rules ⬇️ www.biorxiv.org/content/10.1...
15419
Reposted by Ian Driver
Waggoner Lab @labwaggoner.bsky.social · 22/09/2025
@biorxivpreprint.bsky.social Bispecific Targeting of CHI3L1 and PD-1/PD-L1 Axis as a Novel Therapeutic Strategy for Idiopathic Pulmonary Fibrosis www.biorxiv.org/content/10.1...
021
Reposted by Ian Driver
Waggoner Lab @labwaggoner.bsky.social · 20/09/2025
Inflammation-inducible IL-1-targeted therapy using an rAAV vector as a long-lasting, pathophysiologic treatment for chronic inflammatory diseases @moltherapy.bsky.social @umasschan.bsky.social www.cell.com/molecular-th...
091
Reposted by Ian Driver
bioRxivpreprint @biorxivpreprint.bsky.social · 19/09/2025
Metabolomic, Lipidomic, and Enterohormone Changes in the Progression from MASLD to MASH www.biorxiv.org/content/10.1101/202…
022
Reposted by Ian Driver
bioRxivpreprint @biorxivpreprint.bsky.social · 06/09/2025
Nutrient-Sensing Nuclear Receptor PPARα Controls Liver Proteostasis www.biorxiv.org/content/10.1101/202…
011
Reposted by Ian Driver
Waggoner Lab @labwaggoner.bsky.social · 04/09/2025
PICALM Alzheimer’s risk allele causes aberrant lipid droplets in microglia @nature.com www.nature.com/articles/s41...
052
Reposted by Ian Driver
bioRxiv Bioinfo @biorxiv-bioinfo.bsky.social · 03/09/2025
scAgeClock: a single-cell transcriptome based human aging clock model using gated multi-head attention neural networks www.biorxiv.org/content/10.1101/202…
0104
Reposted by Ian Driver
Waggoner Lab @labwaggoner.bsky.social · 29/08/2025
Single-cell atlas of human liver and blood immune cells across fatty liver disease stages reveals distinct signatures linked to liver dysfunction and fibrogenesis @natimmunol.nature.com www.nature.com/articles/s41...
092
Reposted by Ian Driver
Waggoner Lab @labwaggoner.bsky.social · 28/08/2025
Targeting acid ceramidase enhances collagen degradation by hepatic stellate cells via PKCα and MMP-1, uncovering a pathway to reverse liver fibrosis www.science.org/doi/10.1126/...
041