Benjamin Vincent

A Unifying Grammar for DNA Variants A unified framework and toolkit for standardizing DNA variants at scale. Part 1 of the neoantigen discovery with long-read sequencing series.

When @benjamingvincent.bsky.social and I started PIRL one of our hypotheses was that neoantigen vaccines mostly don’t work bc SNVs create low value (self-like) targets. What other mutational antigens could we use instead? @jslee.bsky.social been working on making those legible…

It’s time: @benjamingvincent.bsky.social and I are going to do a cancer immunotherapy / cool research / accelerating future medicine podcast.

Any tips for getting started?

After next round of the bake-off, going to pivot to tumor antigens (once we get some mass spec validation of which ones are real).

For now, read the blog post by @benjamingvincent.bsky.social: pirlblog.substack.com/p/an-experim...

Three Axes of Innovation in Personalized Cancer Vaccines The personalized cancer vaccine field had its first true clinical success with the KEYNOTE-942 study, where patients with resected melanoma were treated with Merck’s PD-1 inhibitor pembrolizumab (Keyt...

Are we on the cusp of curing cancer w/ mRNA vaccines? …not quite.

“…predicted tumor antigens are not actually presented by tumor…antigens included in personalized tumor vaccines are minimally immunogenic”

@benjamingvincent.bsky.social lays out a path forward

open.substack.com/pub/pirlblog...

(x-post from the bad place)

I'm going to keep talking about this idea until @benjamingvincent.bsky.social & I do it or someone else successfully does it for us (it just needs to exist):

Many cancers express reproductive proteins (aka "cancer/testis antigens"), often at very high levels

1/n

There will likely be a narrowing of lanes around tumor cell-specific targeting though, which you need for your pMHC binders, antibodies or similar, and nucleic acid therapies … same will be true of the cancer-eating nanobots in the further future …

Whoever gets the good training data first wins

It may be a #NewYear, but our resolution remains the same: through our fiercely patient-centric approach, we are still working tirelessly to create greatly needed research infrastructure to improve the outcomes and care for #RareCancer patients. 💙💚

NEW PREPRINT

A detailed overview of 32 popular predictive performance metrics for prediction models

arxiv.org/abs/2412.10288

A tweet: Alex Blechman @AlexBlechman Sci-Fi Author: In my book I invented the Torment Nexus as a cautionary tale Tech Company: At long last, we have created the Torment Nexus from classic sci-fi novel Don't Create The Torment Nexus 5:49 PM Nov 8, 2021 Twitter Web App

I still think about this at least once a week.

“NIH produces an astounding return on investment to the American taxpayer. In Fiscal Year 2022, NIH research funding supported 568,585 jobs and generated $96.84 billion in economic activity — that’s $2.64 of economic activity for every $1 of research funding.”
www.researchamerica.org/2023-oped-us...

T cell receptor–centric perspective to multimodal single-cell data analysis T cell receptor–centric perspective improves single-cell analysis.

Our massive effort to recontextualize #single #T-cell data is now out in Science Advances! We developed a new system, STEGO, to do a TCR-first analysis, and reanalyzed 12 studies and more than 500 000 individual T-cells.
Supported by @cziscience.bsky.social !
doi.org/10.1126/scia...

Yes - it’s awesome how science builds on itself unexpectedly

Your periodic reminder: read the 8th Day of Creation

"The basic principles of the genetic code & the molecular biology of prokaryotes were established in the years between (1953 & 1973) as documented in extraordinary detail in 1979 by Horace Freeland Judson in the “Eighth Day of Creation”

Rachel Thomas, PhD - AI’s Quest to Predict T Cell Binding– The Holy Grail of Immunology an AI researcher going back to school for immunology

Such an interesting read from @math-rachel.bsky.social who has a remarkable gift for making complex topics (like AI ∩ immunology) accessible as well as engaging -> rachel.fast.ai/posts/2024-0... #biosky #medsky

meme paper summary showing a fork in the road for the tcr:pmhc specificity problem: 1. pointing to binary classification 2. pointing towards gen AI

My first Skeetorial!

💻🧬TCR-TRANSLATE - A new framework for thinking about the TCR:pMHC specificity problem.

TLDR:
We pretrained LLMs on ~8M TCR & pMHC seqs
Finetuned on sparse pMHC->TCR pair data
Validated CDR3b sequences to unseen antigens
>> random performance on IMMREP2023 "private" antigens

Inspired by @rodrahimi.bsky.social 's noble efforts linking immunologists together over here, let's try and get more specifically TCR/BCR/AIRR/repertoire chat going with this starter park
go.bsky.app/Gi2dcGd
Drop me a beep to get added or suggest others!

Is Discover feed the best default?

Didn't see this one but now I wish I had

Your regular reminder that we, your professors, LOVE IT when you email us after the fact to let us know what you're up to. Truly one of the most gratifying parts of this work.

I can talk to them if they're interested, but essentially you have to learn a bunch of biology and ideally get deep into assay details to become a useful critical thinker

A real moonshot project in biomedical research would be a decisive legal, regulatory and IT effort to really make data access work for researchers across the world. It's not as sexy as AI, CRISPR, or ginormous sequencing projects, but more transformative and better bang for the buck.

I am very excited about guest editing a special issue in Genome Research (@genomeresearch.bsky.social )
about long reads DNA Sequencing Applications with Ana Conesa & Alexander Hoischen. This is such a timely and fast moving topic about all its applications, innovations etc. bit.ly/grcallforpap...

Huge team effort by a bunch of different people in PIRL to finish a large mouse vaccine experiment, hopefully now we get to learn something about how mRNA compares with different peptide + adjuvant vaccines.

Welcome to our Bluesky account! Follow us to see updates on what’s latest in #massspec and the accomplishments of the Baker lab! 🧪

Trainees - This is a reminder that your PIs want to see ALL the data. We want to see the experiments that "worked" and the experiments that "didn't work". You can learn a lot from results that do not fit your hypothesis. Resist the urge to curate what you show your advisor. 🧪

Congratulations!