←
Dylan Webster134 posts
Dylan Webster avatar

Dylan Webster

@DylanPWebster

Software product @10xGenomics. 🧬⚾🥋🎸

◷ Joined January 2012▧ 28 Photos and videos
Dylan Webster avatar
⟳ Dylan Webster retweeted
Dylan Webster@DylanPWebster·Nov 3

In the notebooks, you see the starting prompt and analysis plan. Our approach for reasoning transparency is our "explain the code" feature in the app itself which you can always go back to. We've gotten feedback that this is great for folks learning analysis steps.

Post attachmentPost attachment
10x Genomics avatar
⟳ Dylan Webster retweeted
10x Genomics@10xGenomics·Nov 2

Introducing Sentira Single Cell, a new agentic analysis platform. Explore Chromium data with natural language, refine analyses in editable, reproducible notebooks and stay in control from scientific questions to experimental conclusions: bit.ly/47PCu18

Post attachmentPost attachmentPost attachment
Dylan Webster avatar
Dylan Webster@DylanPWebster·Nov 2

We released Sentira Single Cell today. It's an agentic analysis application for single cell data with scaled compute in 10x Cloud. Check it out and let us know what you think: 10xgen.com/sentira Congrats to the entire team for making it happen.

1:05
Anshul Kundaje avatar
⟳ Dylan Webster retweeted
Anshul Kundaje@anshulkundaje·Oct 31

'Almost 3,500 GWAS signals across 38 relevant traits colocalize with sn-eQTL/caQTL. These signals typically colocalize with caQTL and not eQTL, highlighting the importance of population-scale chromatin profiling for GWAS functional studies." Been saying this for years ....

We are gradually building single-cell eQTL resources across most human tissues. A new resource now adds skeletal muscle to the picture, leveraging snRNA-seq and snATAC-seq data from 287 individuals.

Post attachment
Dylan Webster avatar
Dylan Webster@DylanPWebster·Oct 30

Super cool. Seems like a lot of related recent work is pointing to a vision of amazing things to come unraveling eQTLs. These are the building blocks that make AI x bio promises tangible.

What kind of data do we need for DNA foundation models? At @RelationRx, we have developed a new protocol for prime-editing perturb-seq to help generate training data. biorxiv.org/content/10.64898… CC @JensBagerChris1 , @j_opzoomer (other authors not on twitter!) Thread below..

Post attachment
10x Genomics avatar
⟳ Dylan Webster retweeted
10x Genomics@10xGenomics·Sep 29

Announcing Sentira, a new cloud analysis platform designed to accelerate the path from complex biological data to experimental conclusions, starting with Sentira Spatial Analysis and @NVIDIA accelerated computing. Now available to Early Access customers: bit.ly/4rA7mvq

Post attachment
Dylan Webster avatar
Dylan Webster@DylanPWebster·Mar 18

Awesome to see this out in the world! Very much looking forward to continuing our collaboration with Samuel Alber (@EggsAndSam7) and the @james_y_zou lab. More to come!

Exciting new work on AI agents in biology from @james_y_zou and @EggsAndSam7 in @naturemethods! As biology shifts from data generation to interpretation, AI agents can help with hypothesis generation. Excited to keep collaborating with the Zou Lab: nature.com/articles/s41592-0…

Post attachmentPost attachment
Dylan Webster avatar
Dylan Webster@DylanPWebster·Mar 17

. @10xGenomics "GEM-X Flex generates reliable, comprehensive transcriptomic data from both fixed tissue and clinical biopsies. By overcoming some of the limitations of fresh and frozen tissue analysis, this protocol offers a robust solution for the broad implementation of scRNA-seq in clinical trials." 👀

Performance of the 10X Genomics Flex Single-Cell Sequencing Assay and its Application to Overcome Challenges in Clinical Trial Samples biorxiv.org/content/10.64898… #biorxiv_genomic

Rahul Satija avatar
⟳ Dylan Webster retweeted
Rahul Satija@satijalab·Feb 17

Excited to share VIPerturb-seq! New tech from my lab which aims to improve the cost, data quality, and efficiency of single-cell CRISPR screens so that they are accessible to any lab - even at genome-wide scale Preprint and 🧵 (1/): biorxiv.org/content/10.64898…

Dylan Webster avatar
Dylan Webster@DylanPWebster·Feb 16

Another point.☝️To move functional genomics into a new era of high-resolution screening for clinical cohorts, we need more than just scale—we need robustness. Recent methods like VIPerturb-seq and ProPer-seq brilliantly demonstrate the advances of @10xGenomics Flex Apex in throughput and cost efficiency with cell lines. While cell lines are excellent for technology development and some modeling applications, the true barrier to large-scale, hypothesis-driven screening is the logistical complexity of handling real-world samples. For these samples, the value of a screen is defined by its reproducibility and its ability to interrogate primary human biology. The 10x Genomics Flex platform is emerging as the standard for these high-value datasets, specifically due to its probe-based fixation workflow. This architecture solves the critical “sampling vs. processing” bottleneck, enabling the standardization required for large cohorts. These recent studies illustrate why this workflow is essential for ensuring ROI on large-scale screens: 1. Standardization of Primary Cohorts (Zhu et al.) Conducting genome-scale screens in primary human cells (such as CD4+ T cells) has historically been limited by donor variability and batch effects. The Flex workflow allows for fixation at the point of collection, enabling teams to “bank and batch” samples. This effectively eliminates technical noise, ensuring that the variance observed is biological. biorxiv.org/content/10.64898… 2. Multi-Modal Fidelity in Intact Tissue (Saunders et al.) Validating targets often requires more than just sequencing—it requires phenotypic context. Saunders et al. utilized this workflow to perform pooled genetic screens directly in intact mammalian tissue. The compatibility of Flex chemistry with fixed, permeabilized samples allows researchers to layer complex readouts (like imaging) with transcriptomics, capturing high-resolution data from complex tissue without the degradation issues common in traditional dissociation protocols. cell.com/cell/fulltext/S0092… For R&D organizations, this represents a shift in operational capability. We can now design the highest resolution functional screens around the biology of the patient cohort, rather than the constraints of the assay. #drugdiscovery #functionalgenomics #biotech #precisionmedicine #10xGenomics #singlecell

And another one. ☝️ Last week: ProPer-seq. This week: VIPerturb-seq from the Satija Lab. Paper after paper is proving @10xGenomics Flex is the definitive engine for genome-scale perturbation screens. The new stats are wild: 🚀 440,000 cells per lane 🚀 50x throughput boost 🚀 65% more genes detected/cell Routine genome-scale screens are here. #singlecell #genomics #CRISPR #biotech #10xGenomics #virtualcell

Post attachment
Robert Stock avatar
⟳ Dylan Webster retweeted
Robert Stock@RobertStock6·Feb 16

Of course that’s your contention. You’re a first-year analytics intern. You just got finished reading some Sabermetric historian—Bill James, probably. You’re gonna be convinced of that ’til next month when you get to the latest Pitching+ update, and then you’re gonna be talkin’ about how SIERA is the only "true" measure of skill because it accounts for the complexity of the batted ball profile to CSW rates. You have no thoughts of your own. You’re just looking at a movement plot and trying to sound smart. You’re gonna realize in a couple of years that you spent all that time calculating "Expected FIP" based on a "Skill-Interactive" ghost, when you could’ve just looked at the K/BB ratio and seen the whole picture. The sad thing is, in about five years, you’re gonna realize you’re not that bright. You’re gonna be talkin’ about Skenes’ "Expected ERA" while the real experts are just looking at a box score and realizing that if a guy doesn't walk people and he misses bats, he's gonna be fine. You dropped $100,000 on a Data Science degree to learn how to run a regression on a pitcher's "luck," when you could’ve gotten the same predictive value for five cents in late charges by looking at a K/BB leader board at the public library.

Post attachment