@boyuan_data

Senior Data Scientist at Merck AI. Ex-Googler at DeepMind. PhD in biological data science at Harvard with Chris Sander and Aviv Regev.

 Cambridge, MA
Joined September 2020
  • Tweets55
  • Following186
  • Followers231
2Media
Thrilled to share our new paper describing a novel form of neural computation: single cells drastically change their tuning in <20ms, shifting from a code for face detection to one that enables fine face discrimination. Neural dynamics are much richer than previously conceived!
Doris Tsao
@doristsao7 Dec 2023
I am beyond excited to share a new paper “Rapid, concerted switching of the neural code in inferotemporal cortex,” led by @Yuelin_Shi and Dasheng Bi. We describe a new form of neural computation in which single neurons rapidly change their tuning to temporally multiplex different functions. In < 20 ms, cells switch from a code optimized for face detection to a code optimized for face discrimination. I think this discovery fundamentally changes our current concept of neural representation. (1/N) biorxiv.org/content/10.1101/…
10
50
6,722
I am extremely excited to announce our next Q&A with the one and only @deboramarks. Come participate in our live Q&A on Slack on September 12 at 1:30pm PST/4:30pm EST (as always, you can find this on our calendar). This is a great opportunity to chat all things ML + bio
1
6
1
30
The Spotlight: A General Method for Discovering Systematic Errors in Deep Learning Models deepai.org/publication/the-s… by Greg d'Eon et al. #NeuralNetwork #Classifier

DeepAI

Artificially intelligent tools for naturally creative humans.

deepai.org
2
6
Interesting interview with Aviv Regev about her move to @genentech: How to deliver better medicine? 1. human biology, 2. high-res methods, 3. modalities, and 4. computation & maths, the latter for "predictive, generative and interpretable models". nature.com/articles/d41573-0…

Aviv Regev

Nature Reviews Drug Discovery - Aviv Regev, head of Genentech Research and Early Development (gRED), discusses how advances in human biology, massively parallel high-resolution methods, and...

nature.com
3
35
1
138
Excited to share our #preprint on inCITE-seq for measuring nuclear protein levels + transcriptome at single cell! Like CITE-seq for cell-surface proteins, we use DNA-conjugated antibodies to measure proteins, but now inside the nucleus + transcriptome. biorxiv.org/content/10.1101/…

Simultaneous single cell measurements of intranuclear proteins and gene expression

Identifying gene regulatory targets of nuclear proteins in tissues remains a challenge. Here we describe intranuclear Cellular Indexing of Transcriptomes and Epitopes (inCITE-seq), a scalable method...

biorxiv.org
8
113
18
423
Allon Klein of @Harvard and #NASmember Aviv Regev of @genentech @broadinstitute @HHMInews, are the inaugural James Prize in Science and Technology Integration recipients! #NASaward #RNA Read more: ow.ly/GBu050DaVy0
1
19
12
144
My thoughts on AlphaFold2: moalquraishi.wordpress.com/2…

AlphaFold2 @ CASP14: “It feels like one’s child has left home.”

The past week was a momentous occasion for protein structure prediction, structural biology at large, and in due time, may prove to be so for the whole of life sciences. CASP14, the conference for …

moalquraishi.wordpress.com
25
222
111
985
Interesting idea! Will give this a try with our next paper!
This Post is from an account that no longer exists. Learn more
12
23
Our paper on speeding up NeuralODE (yes, you do not need to integrate z backwards in time! just interpolate it, and this is stable) got accepted to #NeurIPS2020 arxiv.org/abs/2003.05271

Interpolation Technique to Speed Up Gradients Propagation in Neural ODEs

We propose a simple interpolation-based method for the efficient approximation of gradients in neural ODE models. We compare it with the reverse dynamic method (known in the literature as "adjoint...

arxiv.org
2
10
1
55
Our review (w/ @adamgayoso & Nir @YosefLab) on deep generative models (DGMs) for molecular biology is now available in open-access in @MolSystBiol. I briefly describe the content in the tweets below. 1/n embopress.org/doi/full/10.15…

Enhancing scientific discoveries in molecular biology with deep generative models

Molecular Systems Biology - Generative models provide a well‐established statistical framework for evaluating uncertainty and deriving conclusions from large data sets especially in the...

link.springer.com
2
22
2
79
One of the best (…) self-referential tweets you could imagine :(
Simon Donner@simondonner22 Sep 2020
Uh, we did it.
8
25
Let's try inverting the colors... (h/t @KnabeWolf)
115
2,090
200
21,539
.@Reuters shows how organic carbon is released into the atmosphere during the fires. The smoke contains a substantial portion of fine particulate matter known by the particles’ size as PM2.5, which can have a major impact on people’s health 2/7
16
410
59
823
Replying to @madsommars
Saw this a couple weeks ago and liked it SO much that I incorporated some similar language to my email signature (I'm guilty of finding myself doing work stuff at unspeakable hours)
19
629
71
10,141
“People have felt powerless in this pandemic. A course like this can help people feel like they have the tools to do something about it,” says Young lab graduate student Lena Afeyan, the TA for the course “COVID-19, SARS-CoV-2 & the Pandemic.” Learn more: ow.ly/i7Sj50BlVp0
Whitehead Institute@WhiteheadInst9 Sep 2020
A new MIT course led by Whitehead Institute Member Richard Young and @ragoninstitute’s Facundo Batista brings together top COVID-19 experts to educate students on up-to-date pandemic science. Learn what the course offers and how to access it for free: ow.ly/i7Sj50BlVp0
Screenshot of the COVID-19 course showing the instructors and TA

ALT Screenshot of the COVID-19 course showing the instructors and TA

3
5
New synthetic mammalian circuits that allow cells to sense & control their own pop size: biorxiv.org/content/10.1101/…, Led by brilliant @ma_yitong and @markwbudde w/Michaëlle Mayalu, Junqin Zhu & Richard Murray.

Synthetic mammalian signaling circuits for robust cell population control

In multicellular organisms, cells actively sense, respond to, and control their own population density. Synthetic mammalian quorum sensing circuits could provide insight into principles of population...

biorxiv.org
2
27
2
121
TopoMap: A 0-dimensional Homology Preserving Projection of High-Dimensional Data deepai.org/publication/topom… by Harish Doraiswamy et al. #HighDimensionalData #Outlier

TopoMap: A 0-dimensional Homology Preserving Projection of High-Dimensional Data

09/03/20 - Multidimensional Projection is a fundamental tool for high-dimensional data analytics and visualization. With very few exceptions,...

deepai.org
7
20
Understanding the wiring evolution in differentiable neural architecture search deepai.org/publication/under… by Sirui Xie et al. #ComputerScience #Learning
2
7
I skimmed all papers related to reasoning from NeurIPS, ICML, and ICLR since 1987. I made a shortlist with ~2700 papers. It's work-in-progress, but I thought the raw list could be useful if you are exploring research directions. docs.google.com/spreadsheets…

ML Reasoning Reading list

docs.google.com
11
68
5
333
DeepVariant 1.0 is released! Notably: "Compared to DeepVariant v0.10, these changes reduce Illumina WGS errors by 24%, exome errors by 19%, and PacBio errors by 52%." github.com/google/deepvarian… Plus a new hybrid PacBio+Illumina model: github.com/google/deepvarian…

Release DeepVariant 1.0.0 · google/deepvariant

DeepVariant v1.0 releases new features and accuracy improvements sufficiently substantial to indicate a major version of v1.0. Compared to DeepVariant v0.10, these changes reduce Illumina WGS error...

github.com
5
58
211
Load more