Blindspot Bio

Processing-bias correction with DEBIAS-M improves cross-study generalization of microbiome-based prediction models - Nature Microbiology DEBIAS-M corrects technical variability in microbiome data in a manner both interpretable and suitable for machine learning. In extensive benchmarks, DEBIAS-M facilitates robust analyses that generali...

Happy to share DEBIAS-M, our new method for domain adaptation and bias correction in #microbiome data.🧬🖥️

Microbiome data is very variable, with substantial study- and batch-effects. DEBIAS-M corrects these, enabling robust and generalizable analyses.
A quick thread:
www.nature.com/articles/s41...

CTP-1.0 CTP guidelines 1.0 Definition and Operational Details of the Crowd-to-Paper Model This document, called “CTP-1.0”, is a template that includes the guiding principles of projects based on the crowd-to-...

Would you open your paper to a modular open-collaboration model?

It was game-changing for #tidyomics

If you love #crowdResearch and #CommunityBuilding have your say on the #crowdToPaper model guidelines

Comment on the shared doc: docs.google.com/document/d/1...

Please share 🙏

It took 35 years for the first human Mendelian disease to be described Mendel first described his laws of genetic inheritance in 1865. They were promptly ignored for 35 years.

Mendel first described his laws of genetic inheritance in 1865. They were promptly ignored for 35 years. 🧬 🧪 #HistSci

The benefits of preregistration and Registered Reports Practices that introduce systematic bias are common in most scientific disciplines, including toxicology. Selective reporting of results and publication bias are two of the most prevalent sources o...

If you would like to read what I hope (and tried) to make the best (=most conceptually rigorous) review of preregistration and Registered Reports currently in the literature, see our recent paper: www.tandfonline.com/doi/full/10....

E. coli prepares for starvation by dramatically remodeling its proteome in the first hours after loss of nutrients It is widely believed that due to nutrient limitations in natural environments, bacteria spend most of their life in non-growing states. However, very little is known about how bacteria change their p...

What do bacterial cells do when they run out of nutrients? Although most bacterial studies focus on cells in exponentially growing states, in the wild bacteria likely spend most of their time slowly starving to death. 1/n
www.biorxiv.org/content/10.1...

.doom.d/snippets/python-mode/uvbang at 3f53f48071035fdc498a5cf36fff31e51b537367 · edmundmiller/.doom.d My private module for Doom Emacs. Contribute to edmundmiller/.doom.d development by creating an account on GitHub.

Made a snippet for a uv run shebang because I got tired of hunting it down everytime I wanted to add it to a new script:
github.com/edmundmiller...

Spatial transcriptomic imaging of an intact organism using volumetric DNA microscopy - Nature Biotechnology Volumetric DNA microscopy captures spatial information of RNA within an intact organism.

Spatial transcriptomic imaging of an intact organism. Mind blowing

www.nature.com/articles/s41...

The tumour histopathology “glossary” for AI developers The applications of artificial intelligence (AI) and deep learning (DL) are leading to significant advances in cancer research, particularly in analysing histopathology images for prognostic and treat...

From @plos.org #Computational #Biology | The tumour histopathology "glossary" for AI developers | #Bioinformatics #Education #AI | 🧬 🖥️ 🧪
see ⬇️
journals.plos.org/ploscompbiol...

The SCGE has recently released a Gene Therapy Clinical Trial Browser. This publicly accessible, free database was created for the benefit of users seeking information on gene therapy development. 🧬🧪

Check out the browser at: scge.mcw.edu/platform/dat...

#clinicaltrials #genetherapy #research

Integration of 168,000 samples reveals global patterns of the human gut microbiome The Human Microbiome Compendium is a unified database of publicly available human gut microbiome 16S samples, built with the integrated data from hundreds of independent projects. The compendium is us...

Today we report a new compendium of human gut microbiomes with >168,000 samples

By analyzing this massive dataset, we discovered distinct microbiome patterns across the globe, and show we can predict where a person lives just from their gut bacteria

Now out in Cell:

www.cell.com/cell/fulltex...

Microbial ecosystems and ecological driving forces in the deepest ocean sediments Metagenomic sequencing of 1,648 sediment samples from 6–11 km water depths, including the Mariana Trench, highlights the hadal microbial ecosystem and its environmental driving forces.

Microbial ecosystems and ecological driving forces in the deepest ocean sediments www.cell.com/cell/fulltex... #jcampubs

Evidence Aggregator: AI reasoning applied to rare disease diagnostics Retrieving, reviewing, and synthesizing technical information can be time-consuming and challenging, particularly when requiring specialized expertise, as is the case of variant assessment for rare di...

New preprint! We worked with @msftresearch.bsky.social and @broadinstitute.org to see whether large language models (LLMs) can be useful to variant scientists in deciding whether genetic variants seen in a patient are responsible for their disease. tl;dr yes they can: www.biorxiv.org/content/10.1...

Silhouettes of infant, mouse, chick, and bacteria.

Interested in the influence of the infant microbiome on health? Quickly interrogating non-model organisms using omic approaches?

A new publication, describing genome-scale resources in bifidobacteria, is now available online: doi.org/10.1016/j.ce...

#microbiome #microsky

Biased sampling confounds machine learning prediction of antimicrobial resistance Antimicrobial resistance (AMR) poses a growing threat to human health. Increasingly, genome sequencing is being applied for the surveillance of bacterial pathogens, producing a wealth of data to train...

Interested in predicting AMR in bacteria? We show ignoring phylogenetic structure in genome collections leads to overly optimistic evaluations of machine learning methods for AMR prediction. With @nwheeler443.bsky.social and former PhD student Yanying Yu. 🦠🧫🧬🖥️🧪

www.biorxiv.org/content/10.1...

Command-line interface: running "ska" and creating a phylogenetic tree

🌲 New sandbox.bio tutorial: Learn to build phylogenetic trees from closely related genomes using ska (the tool, not the music) 🖥️ 🧬

sandbox.bio/tutorials/sk...

Taking Stock of the Known and Unknown Microbial Space - Biosciences Area Using publicly available genome sequence data generated over the past three decades, JGI researchers assess the known fraction of microbial diversity.

In Science Advances: "We took a deep dive into over 1.8 million bacterial and archaeal genomes to see how much of their diversity we’ve actually captured. Turns out that despite all the genomes we’ve sequenced, we’ve only scratched the surface." -Dongying Wu

🖥️🧬🦠

biosciences.lbl.gov/2025/01/17/t...

Does anyone have any favorite reviews of autoencoders or VAEs for genomics data?

Review: Artificial intelligence in drug development https://www.nature.com/articles/s41591-024-03434-4 (read free: https://rdcu.be/d63NO) 🧬🖥️🧪

Cancer burden is shifting from men to women and old to young Cancer deaths overall are falling, but the burden is shifting toward younger adults and women, the American Cancer Society reports.

Deaths from cancer continue to fall in the United States, the American Cancer Society reported, but within that encouraging trend is a disturbing shift in the cancer burden from older to younger adults and from men to women.

GENESPACE tracks regions of interest and gene copy number variation across multiple genomes GENESPACE lets users track related chromosomal sequences across multiple reference genomes.

But even this can be too much, especially since many genomic analyses often focus on genes. One great example for analyzing and visualizing how genes vary in a pangenome is GENESPACE (this came up over and over again at #PAG32 | #PAG2025) elifesciences.org/articles/785...

So long San Diego! Here’s my opinionated meeting report from #PAG2025 | #PAG32. What else did I miss???

What’s currently the best computational tool for predicting mammalian promoters from DNA sequence alone?

Very excited to announce that the single cell/nuc. RNA/ATAC/multi-ome resource from ENCODE4 is now officially public. This includes raw data, processed data, annotations and pseudobulk products. Covers many human & mouse tissues. 1/

www.encodeproject.org/single-cell/...

Visidata GFF Pplugin Recorded by emiller

Something that had been on my wishlist for a long time was a #Visidata plugin for common tabular #bioinformatics formats. I got by with just telling Visidata they were TSVs.

Finally got around to writing a GFF plugin
asciinema.org/a/Nt0AlgvxsX...
🧬🖥️

Top 20 most studied bacteria shown based on the number of PubMed articles.

Chart showing 50% of papers focus on just ten microbial species and 74% of microbial species (out of 43,409) are not featured in individual papers.

Popular microbes hogs the spotlight while the majority remain in the shadows. www.nature.com/articles/d41...

Large Language Models for Bioinformatics With the rapid advancements in large language model (LLM) technology and the emergence of bioinformatics-specific language models (BioLMs), there is a growing need for a comprehensive analysis of the ...

Large Language Models for Bioinformatics arxiv.org/abs/2501.06271

5,562 proteins were found to have tissue-specific enrichment across 32 human tissues.

The majority of them (3,088) are enriched at the protein but not RNA level.

The brain hosts the largest number of proteins that are only enriched at the protein level but not RNA level.

Computational advances in discovering cryptic pockets for drug discovery A number of promising therapeutic target proteins have been considered “undruggable” due to the lack of well-defined ligandable pockets. Substantial r…

Cryptic pockets are hidden binding sites in proteins that are not visible in their standard ("apo") structure and only stable in the presence of the right ligand.

Here, Bemelmans et al. review how scientists use computer algorithms to detect them.
www.sciencedirect.com/science/arti...

🧪 #CompChem

If you hate statistics like I do, then you'll love my free lectures. Putting science before statistics, 20 lectures from basics of inference & causal modeling to multilevel models & dynamic state space models. It's all free, made with love and sympathy. 🧪 #stats www.youtube.com/playlist?lis...