Alex Rubinsteyn
@alexr.bsky.social
personalized cancer immunotherapy = genomics + immunology + machine learning + oncology
(pirl.unc.edu)
(pirl.unc.edu)
I feel like I have the opposite orientation to most people on here wrt my writing getting used to train genAI models. I worry about the paywalled articles not getting included in science reasoning training data
(And feel some shame about anything I wrote being paywalled in the first place)
(And feel some shame about anything I wrote being paywalled in the first place)
November 27, 2025 at 2:37 PM
I feel like I have the opposite orientation to most people on here wrt my writing getting used to train genAI models. I worry about the paywalled articles not getting included in science reasoning training data
(And feel some shame about anything I wrote being paywalled in the first place)
(And feel some shame about anything I wrote being paywalled in the first place)
OK, trying my first "design an immunology experiment" eval query on ChatGPT 5.1 {Pro, Thinking} / Gemini 3 / Claude 4.5 (Thinking) / Edison {Kosmos, Literature Search} / AI2 Asta / Elicit / Perplexity
Any others I should include?
(Tried to use Causaly but never got an email)
Any others I should include?
(Tried to use Causaly but never got an email)
November 25, 2025 at 8:14 PM
OK, trying my first "design an immunology experiment" eval query on ChatGPT 5.1 {Pro, Thinking} / Gemini 3 / Claude 4.5 (Thinking) / Edison {Kosmos, Literature Search} / AI2 Asta / Elicit / Perplexity
Any others I should include?
(Tried to use Causaly but never got an email)
Any others I should include?
(Tried to use Causaly but never got an email)
Have any immunology labs gotten TScan working in their own hands?
November 25, 2025 at 8:06 PM
Have any immunology labs gotten TScan working in their own hands?
Reposted by Alex Rubinsteyn
Ok; mim (github.com/COMBINE-lab/...) preprint submitted! Excited for folks to see it and share thoughts. The key takeaway; mim allows the quick, one-time, building of a small auxiliary index that then allows scaling gzipped FASTQ parsing linearly in # of threads. 1/2
GitHub - COMBINE-lab/mim: A small, auxiliary index to massively improve parallel fastq parsing
A small, auxiliary index to massively improve parallel fastq parsing - COMBINE-lab/mim
github.com
November 25, 2025 at 2:13 PM
Ok; mim (github.com/COMBINE-lab/...) preprint submitted! Excited for folks to see it and share thoughts. The key takeaway; mim allows the quick, one-time, building of a small auxiliary index that then allows scaling gzipped FASTQ parsing linearly in # of threads. 1/2
Impressions of first real bio deep dive with Gemini 3 — it seems almost on par with ChatGPT 5 Pro, can’t tell if it’s a little worse or I’m not used to it — but definitely grabbing right eg lipid ratios from correct papers but! The suggested next steps are way worse. Weird
November 25, 2025 at 2:10 AM
Impressions of first real bio deep dive with Gemini 3 — it seems almost on par with ChatGPT 5 Pro, can’t tell if it’s a little worse or I’m not used to it — but definitely grabbing right eg lipid ratios from correct papers but! The suggested next steps are way worse. Weird
Got stuck on a few different experiment design queries with ChatGPT 5 Pro
Finally got the Gemini app and tried Gemini 3 and…it perfectly answered my question with every relevant experimental detail across a few different papers/patents
Finally got the Gemini app and tried Gemini 3 and…it perfectly answered my question with every relevant experimental detail across a few different papers/patents
November 25, 2025 at 12:54 AM
Got stuck on a few different experiment design queries with ChatGPT 5 Pro
Finally got the Gemini app and tried Gemini 3 and…it perfectly answered my question with every relevant experimental detail across a few different papers/patents
Finally got the Gemini app and tried Gemini 3 and…it perfectly answered my question with every relevant experimental detail across a few different papers/patents
Which AI tools have you tried for science (e.g. lit review or experiment brainstorming)?
Which did you find useful?
Which did you find useful?
November 23, 2025 at 11:45 PM
Which AI tools have you tried for science (e.g. lit review or experiment brainstorming)?
Which did you find useful?
Which did you find useful?
Thought of a decent eval for “AI scientist” / “AI science assistant” tools
Other than Kosmos, ChatGPT {Thinking, Pro}, Claude/Opus, and Gemini;
what else is worth trying for an experimental design question that requires integrating across a few different corners of literature?
Other than Kosmos, ChatGPT {Thinking, Pro}, Claude/Opus, and Gemini;
what else is worth trying for an experimental design question that requires integrating across a few different corners of literature?
November 23, 2025 at 3:41 PM
Thought of a decent eval for “AI scientist” / “AI science assistant” tools
Other than Kosmos, ChatGPT {Thinking, Pro}, Claude/Opus, and Gemini;
what else is worth trying for an experimental design question that requires integrating across a few different corners of literature?
Other than Kosmos, ChatGPT {Thinking, Pro}, Claude/Opus, and Gemini;
what else is worth trying for an experimental design question that requires integrating across a few different corners of literature?
ChatGPT is such a delightful superpower for niche bio research
Needed a constrained kind of binder for a constrained subset of surface proteins on a certain subset of lymphocytes and just asking the sky gods while putting my kids to sleep surfaced a dozen good papers
Needed a constrained kind of binder for a constrained subset of surface proteins on a certain subset of lymphocytes and just asking the sky gods while putting my kids to sleep surfaced a dozen good papers
November 20, 2025 at 11:25 PM
ChatGPT is such a delightful superpower for niche bio research
Needed a constrained kind of binder for a constrained subset of surface proteins on a certain subset of lymphocytes and just asking the sky gods while putting my kids to sleep surfaced a dozen good papers
Needed a constrained kind of binder for a constrained subset of surface proteins on a certain subset of lymphocytes and just asking the sky gods while putting my kids to sleep surfaced a dozen good papers
Schools and hospitals scrambling to keep marauding kidnappers at bay, dark shameful stuff going on
November 20, 2025 at 2:22 PM
Schools and hospitals scrambling to keep marauding kidnappers at bay, dark shameful stuff going on
Met Adam Grippen (first author) yesterday and I think his new analysis fixes a lot of these problems and still points to a real but smaller effect
I've been asked about this paper (www.nature.com/articles/s41...) enough times that I finally had to read it.
The basic claim is that mRNA vaccines had huge synergistic benefit with checkpoint blockade against lung cancer melanoma.
tl;dr it's glam-slop & everyone involved should be embarrassed
The basic claim is that mRNA vaccines had huge synergistic benefit with checkpoint blockade against lung cancer melanoma.
tl;dr it's glam-slop & everyone involved should be embarrassed
SARS-CoV-2 mRNA vaccines sensitize tumours to immune checkpoint blockade - Nature
mRNA vaccines targeting SARS-CoV-2 also sensitize tumours to immune checkpoint inhibitors.
www.nature.com
November 19, 2025 at 12:00 AM
Met Adam Grippen (first author) yesterday and I think his new analysis fixes a lot of these problems and still points to a real but smaller effect
Hard to reconcile the claimed high ex vivo CD8+ T-cell responses of FixVac and its imperceptible clinical activity compared with single epitope TCR-Ts
Either immunogenicity data has some huge catch or there’s something important to understand in the discrepancy
Either immunogenicity data has some huge catch or there’s something important to understand in the discrepancy
November 18, 2025 at 11:55 PM
Hard to reconcile the claimed high ex vivo CD8+ T-cell responses of FixVac and its imperceptible clinical activity compared with single epitope TCR-Ts
Either immunogenicity data has some huge catch or there’s something important to understand in the discrepancy
Either immunogenicity data has some huge catch or there’s something important to understand in the discrepancy
Other than neoantigen vaccines, has any other process been in trials which chooses therapeutic sequences computationally & then uses quick commoditized manufacturing (DNA/RNA/peptides)?
What about personalization through an assay (eg neoAg TCR-Ts, bacteriophage bank, n=1 ASOs)?
What about personalization through an assay (eg neoAg TCR-Ts, bacteriophage bank, n=1 ASOs)?
Automated medicine =
molecular profiling as input data for:
computational target selection ->
computational therapeutic design ->
computational safety screening ->
on demand manufacturing using commoditized modalities (DNA, RNA, peptides)
molecular profiling as input data for:
computational target selection ->
computational therapeutic design ->
computational safety screening ->
on demand manufacturing using commoditized modalities (DNA, RNA, peptides)
November 7, 2025 at 8:07 PM
Other than neoantigen vaccines, has any other process been in trials which chooses therapeutic sequences computationally & then uses quick commoditized manufacturing (DNA/RNA/peptides)?
What about personalization through an assay (eg neoAg TCR-Ts, bacteriophage bank, n=1 ASOs)?
What about personalization through an assay (eg neoAg TCR-Ts, bacteriophage bank, n=1 ASOs)?
Automated medicine =
molecular profiling as input data for:
computational target selection ->
computational therapeutic design ->
computational safety screening ->
on demand manufacturing using commoditized modalities (DNA, RNA, peptides)
molecular profiling as input data for:
computational target selection ->
computational therapeutic design ->
computational safety screening ->
on demand manufacturing using commoditized modalities (DNA, RNA, peptides)
November 7, 2025 at 3:47 AM
Automated medicine =
molecular profiling as input data for:
computational target selection ->
computational therapeutic design ->
computational safety screening ->
on demand manufacturing using commoditized modalities (DNA, RNA, peptides)
molecular profiling as input data for:
computational target selection ->
computational therapeutic design ->
computational safety screening ->
on demand manufacturing using commoditized modalities (DNA, RNA, peptides)
Please expand this genre for me:
Ekh Lyuli (by Psoy Korolenko et al & Rajtaraj)
Daloy Politsey (by Rajtaraj but not other versions)
L’chaim Stalin (by Psoy Korolenko et al)
Man with a Hat (by the Klezmatics)
This is all my kids want, not even other songs by the same artists
Ekh Lyuli (by Psoy Korolenko et al & Rajtaraj)
Daloy Politsey (by Rajtaraj but not other versions)
L’chaim Stalin (by Psoy Korolenko et al)
Man with a Hat (by the Klezmatics)
This is all my kids want, not even other songs by the same artists
November 4, 2025 at 1:39 PM
Please expand this genre for me:
Ekh Lyuli (by Psoy Korolenko et al & Rajtaraj)
Daloy Politsey (by Rajtaraj but not other versions)
L’chaim Stalin (by Psoy Korolenko et al)
Man with a Hat (by the Klezmatics)
This is all my kids want, not even other songs by the same artists
Ekh Lyuli (by Psoy Korolenko et al & Rajtaraj)
Daloy Politsey (by Rajtaraj but not other versions)
L’chaim Stalin (by Psoy Korolenko et al)
Man with a Hat (by the Klezmatics)
This is all my kids want, not even other songs by the same artists
What's the funniest dataset to use in an intro ML lecture?
(currently have slides with Onion-or-Not...)
(currently have slides with Onion-or-Not...)
November 4, 2025 at 12:09 PM
What's the funniest dataset to use in an intro ML lecture?
(currently have slides with Onion-or-Not...)
(currently have slides with Onion-or-Not...)
TIL: the guy who originally wrote L’ Chaim Stalin was later executed by Stalin
(still a banger)
(still a banger)
November 1, 2025 at 8:06 PM
TIL: the guy who originally wrote L’ Chaim Stalin was later executed by Stalin
(still a banger)
(still a banger)
Reposted by Alex Rubinsteyn
I love asking the question: how much should be spent on scientific replication? When asked, I have suggested that 1-2% of a funding budget felt right. "Felt right" meaning no evidence basis other than intuition.
@jdworkin.bsky.social offers a much more thoughtful approach.
@jdworkin.bsky.social offers a much more thoughtful approach.
How Much Should We Spend on Scientific Replication? | IFP
A data-driven framework for targeting replication funding where it matters most
ifp.org
October 31, 2025 at 7:40 PM
I love asking the question: how much should be spent on scientific replication? When asked, I have suggested that 1-2% of a funding budget felt right. "Felt right" meaning no evidence basis other than intuition.
@jdworkin.bsky.social offers a much more thoughtful approach.
@jdworkin.bsky.social offers a much more thoughtful approach.
Holy cow, what a difference switching from TCR-TCE to TCR-T seems to make...
Went from "cool that this works at all but only at higher E:T, wonder if the surviving fraction of cells will acquire resistance" to "nuked everything from space at 1:1, leave no cancer cell alive"
Went from "cool that this works at all but only at higher E:T, wonder if the surviving fraction of cells will acquire resistance" to "nuked everything from space at 1:1, leave no cancer cell alive"
October 31, 2025 at 4:16 PM
Holy cow, what a difference switching from TCR-TCE to TCR-T seems to make...
Went from "cool that this works at all but only at higher E:T, wonder if the surviving fraction of cells will acquire resistance" to "nuked everything from space at 1:1, leave no cancer cell alive"
Went from "cool that this works at all but only at higher E:T, wonder if the surviving fraction of cells will acquire resistance" to "nuked everything from space at 1:1, leave no cancer cell alive"
Is the ZSA voyager any good?
(I desperately don't want to configure or customize a keyboard in any way, but do want an ergonomic / joint friendly setup for working from a coffee shop)
(I desperately don't want to configure or customize a keyboard in any way, but do want an ergonomic / joint friendly setup for working from a coffee shop)
October 31, 2025 at 2:23 PM
Is the ZSA voyager any good?
(I desperately don't want to configure or customize a keyboard in any way, but do want an ergonomic / joint friendly setup for working from a coffee shop)
(I desperately don't want to configure or customize a keyboard in any way, but do want an ergonomic / joint friendly setup for working from a coffee shop)
I need a song in the style of L’chaim Stalin….that isn’t about Stalin
youtu.be/MuZDEphRbes
My kids like this song and immediately started singing it
(Taking all recommendations)
youtu.be/MuZDEphRbes
My kids like this song and immediately started singing it
(Taking all recommendations)
L'chaim Stalin
YouTube video by Daniel Kahn & the Painted Bird - Topic
youtu.be
October 31, 2025 at 11:46 AM
I need a song in the style of L’chaim Stalin….that isn’t about Stalin
youtu.be/MuZDEphRbes
My kids like this song and immediately started singing it
(Taking all recommendations)
youtu.be/MuZDEphRbes
My kids like this song and immediately started singing it
(Taking all recommendations)
I've been asked about this paper (www.nature.com/articles/s41...) enough times that I finally had to read it.
The basic claim is that mRNA vaccines had huge synergistic benefit with checkpoint blockade against lung cancer melanoma.
tl;dr it's glam-slop & everyone involved should be embarrassed
The basic claim is that mRNA vaccines had huge synergistic benefit with checkpoint blockade against lung cancer melanoma.
tl;dr it's glam-slop & everyone involved should be embarrassed
SARS-CoV-2 mRNA vaccines sensitize tumours to immune checkpoint blockade - Nature
mRNA vaccines targeting SARS-CoV-2 also sensitize tumours to immune checkpoint inhibitors.
www.nature.com
October 30, 2025 at 3:35 PM
I've been asked about this paper (www.nature.com/articles/s41...) enough times that I finally had to read it.
The basic claim is that mRNA vaccines had huge synergistic benefit with checkpoint blockade against lung cancer melanoma.
tl;dr it's glam-slop & everyone involved should be embarrassed
The basic claim is that mRNA vaccines had huge synergistic benefit with checkpoint blockade against lung cancer melanoma.
tl;dr it's glam-slop & everyone involved should be embarrassed
Good to post this every few months: www.youtube.com/watch?v=wh7d...
Inner Emigration
YouTube video by Daniel Kahn & the Painted Bird - Topic
www.youtube.com
October 24, 2025 at 5:36 PM
Good to post this every few months: www.youtube.com/watch?v=wh7d...