@uthsav.bsky.social
Assistant Professor at JHU CS developing statistical/ML methods for biological applications
The Department of Computer Science is pleased to welcome nine new tenure-track faculty to its ranks this academic year! Featuring @anandbhattad.bsky.social, @uthsav.bsky.social, @gligoric.bsky.social, @murat-kocaoglu.bsky.social, @tiziano.bsky.social, and more:
29.10.2025 13:43 β π 10 π 4 π¬ 0 π 110 new CS professors! π₯³
@anandbhattad.bsky.social @uthsav.bsky.social @gligoric.bsky.social @murat-kocaoglu.bsky.social @tiziano.bsky.social
NSF just cancelled ALL grants to Harvard researchers. Thatβs right - physics, astronomy, bio, CAREER - ALL. Professors wonβt get paid. Postdocs wonβt get paid. PhD students wonβt get paid. This is insane!
If they can do this to Harvard, they can do this to your school.
Wow. "NIH" canceled my co-mentored (with Dave Sulzer) PhD student's F31 funding. His work is on understanding the genetics and neuroscience of language learning disorders. F31 provides no indirect $ to Columbia, just pays his salary. Not that it should matter, but he's an American citizen. W.T.F.
11.03.2025 12:41 β π 521 π 221 π¬ 21 π 17
Computer Science Seminar Series. Tuning Our Algorithmic Amplifiers: Embedding Pro-Social Values Into Online Platforms. March 13, 2025, 228 Malone Hall. Refreshments available 10:30 a.m. Seminar begins 10:45 a.m. Tiziano Piccardi, Stanford University.
CS & BME Seminar Series: Machine Learning for Spatial and Network Biology. March 13, 2025, 12 p.m. 228 Malone Hall. Refreshments available at noon. Seminar begins 12:15 p.m. Uthsav Chitra, Eric and Wendy Schmidt Center at Broad Institute.
2β£ seminars coming up on Thursday with @tiziano.bsky.social and @uthsav.bsky.social! Check them out at bit.ly/3Fv5n7C and bit.ly/3FssGPr
10.03.2025 19:05 β π 2 π 2 π¬ 0 π 0
Finally, if you've gotten this far: this work builds on our earlier work with @congma.bsky.social using ~complex analysis~ (conformal mapping) to model spatial variation in ST data: www.sciencedirect.com/science/arti...
24.01.2025 21:27 β π 0 π 0 π¬ 0 π 0Check out the paper for more details and neat biological applications! For example, modeling spatial gradients β much more accurate SVG identification
Thanks to great collaborators: Brian @hrksrkr.bsky.social Kohei @congma.bsky.social Sereno @braphael.bsky.social
GASTON algorithm: parametrize functions h,d with neural nets and learn from data!
Fun back-story: @braphael.bsky.social and I derived most of this model at the bar near an NCI workshop π₯π
Model implicitly accounts for sparsity by βpoolingβ measurements across locations (x,y) with equal isodepth d(x,y).
These locations look like contours of equal height on an elevation map, hence the βtopographic mapβ analogy.
We prove that f(x,y) = h(d(x,y)), i.e. gene expression f(x,y) is function of a *single* spatial coordinate d(x,y) rather than 2 spatial coordinates x,y
-> Spatial dimensionality reduction! π
We call d(x,y) the "isodepth" - it characterizes spatial gradients β½f_g
We handle sparsity w/ two assumptions:
(1) genes have *shared* gradient directions, i.e. each gradient β½f_g(x,y) is proportional to shared vector field v(x,y)
(Equivalent to Jacobian of f being rank-1 everywhere)
(2) vector field v has no βcurlβ, so v=β½d is gradient of "spatial potential" d
You can view ST data as samples of function f: R^2 β R^G, where f(x,y) is (high-dim) gene expression vector at location (x,y).
Spatial gradients are gradients β½f_g of each component (gene)
Unfortunately, large data sparsity means naive estimation of gradient β½f_g is very noisy π±
GASTON, our method to learn βtopographic mapsβ of gene expression, is out now @naturemethods.bsky.social!
IMO the coolest part is a new model of *spatial gradients in sparse data*.
As is typical for bio papers, itβs buried in Methods, but see below for a quick outline on the math π
A must-read article by @sashagusevposts.bsky.social (hopefully by those whose voice will have an impact). Personally, I'd advocate for a fruit-based distribution system of funds.
theinfinitesimal.substack.com/p/distinguis...
