Pinned Tweet
Essay from Susan Lindquist:
molbiolcell.org/doi/10.1091/…
I’m out. I can’t support a platform intended to misinform and mislead the masses. See you in a better space.
m.youtube.com/watch?v=QDydKw…
145
The root of this network is JJ Thomson. #TeamMassSpec
Nature is surprised academic lineages coalesce if you go back far enough 😂
This is yet another example of something that appears surprising and informative, but it’s actually just a lack of a useful null model being specified.
nature.com/immersive/d41586-…
149
Might anyone have MRM tips or suggestions for improving signal response for more hydrophobic peptides in an isotopic response curve experiment?
1
6
659
For context, this is a result from Experiment #1 of the CPTAC assay validation guidelines: proteomics.cancer.gov/sites/…
1
192
Agar lot-specific inhibition in the plating efficiency of yeast spores and cells academic.oup.com/g3journal/a…
1
88
New paper from my graduate school lab! The Wahls lab used a CRISPR-Cas system to conduct genome editing studies of the ade6 locus and report their findings: unanticipated and undesired genome alterations. Check out the paper to learn more!
mdpi.com/2218-273X/14/8/1016
1
161
Ever Wiki’d a math concept and been confused by the second sentence? I’ve done it so many times, but I found a video series that may help with understanding.
m.youtube.com/watch?v=KCSZ4Q…
73
Retweeting this for my own education.
"Duality" is a deep concept in mathematics, but an intuitive way to think about it is in terms of tables.
When we lay out data in a table, we are familiar with the idea that each row represents a different "entity", and each column represents a different "property". (1/10)
110
Congratulations to the Mayo Clinic on developing a membranous assay to detect 14 membranous antigens in a clinical setting! These kinds of efforts are continuously needed to advance the understanding of kidney disease and provide opportunities for improved care. Excellent work!
Detection of membranous nephropathy (MN) antigens on kidney biopsy.
Laser microdissection and mass spectrometry is a sensitive, specific and reliable one stop method for detection of MN antigens on the kidney bx tissue.
kidney-international.org/art…
Exemplary work from Jeffrey Whiteaker, Amanda Paulovich, and their group.
nature.com/articles/s41597-0…
Picked up a Kingfisher Flex from a liquidation auction and now I feel like I’ve been doing pulldowns wrong my entire career.
1
3
576
Characterization of filament structures from the brains of individuals with FTLD-TDP Type C by Cryo-EM reveals heteromeric amyloid filaments.
Absolutely wild discovery by the @bwfalcon lab! First description of a neurodegenerative disease-associated hetero-amyloid (TDP43+ANXA11): biorxiv.org/content/10.1101/…
141
Excited to talk at #ASMS2024 about developing an MRM for typing membranous nephropathy at @arkanalabs. WOA pm: Clinical Analysis: Applications.
8 years later, this is still my favorite scientific talk. I am still learning from this video. What I appreciate now is their initial work on biomolecular condensates began with studying the nephrin signaling pathway in podocytes. Starts at 17:35.
youtube.com/watch?v=5VWEI04d…
It is an amusing coincidence that the full analysis of ILSVRC2012 was published on December 21st of that year.
# CUDA/C++ origins of Deep Learning
Fun fact many people might have heard about the ImageNet / AlexNet moment of 2012, and the deep learning revolution it started.
en.wikipedia.org/wiki/AlexNe…
What's maybe a bit less known is that the code backing this winning submission to the contest was written from scratch, manually in CUDA/C++ by Alex Krizhevsky. The repo was called cuda-convnet and it was here on Google Code:
code.google.com/archive/p/cu…
I think Google Code was shut down (?), but I found some forks of it on GitHub now, e.g.:
github.com/ulrichstern/cuda-…
This was among the first high-profile applications of CUDA for Deep Learning, and it is the scale that doing so afforded that allowed this network to get such a strong performance in the ImageNet benchmark. Actually this was a fairly sophisticated multi-GPU application too, and e.g. included model-parallelism, where the two parallel convolution streams were split across two GPUs.
You have to also appreciate that at this time in 2012 (~12 years ago), the majority of deep learning was done in Matlab, on CPU, in toy settings, iterating on all kinds of learning algorithms, architectures and optimization ideas. So it was quite novel and unexpected to see Alex, Ilya and Geoff say: forget all the algorithms work, just take a fairly standard ConvNet, make it very big, train it on a big dataset (ImageNet), and just implement the whole thing in CUDA/C++. And it's in this way that deep learning as a field got a big spark. I recall reading through cuda-convnet around that time like... what is this :S
Now of course, there were already hints of a shift in direction towards scaling, e.g. Matlab had its initial support for GPUs, and much of the work in Andrew Ng's lab at Stanford around this time (where I rotated as a 1st year PhD student) was moving in the direction of GPUs for deep learning at scale, among a number of parallel efforts.
But I just thought it was amusing, while writing all this C/C++ code and CUDA kernels, that it feels a bit like coming back around to that moment, to something that looks a bit like cuda-convnet.
155
A glass menagerie of low complexity sequences sciencedirect.com/science/ar…
125
A big year for APOE, recently found in dense deposits and lipid droplets: nature.com/articles/s41586-0…
Talk about burying the lede:
Nice feature article just out in @Nature with focus on our lab's condensate tech (optoDrop, Corelet) & its application in finding treatments for disease. Thrilling to see our spinout Nereid Therapeutics & other companies moving toward the clinic! nature.com/articles/d41586-0…
3
204













