David Earl Hostallero

List of Papers using Deep Learning for Protein Design (by @PeldomZ ) github.com/Peldom/papers_…

📢 Our latest paper is out in @CellCellPress! Led by the multi-talented @XinheZheng, we accelerated AAV-based gene delivery to 2 days & enabled in vivo CRISPR screens with single cell-level readout, a leap from existing capabilities 🤯 How does it work? (1/8)

MLPs are so foundational, but are there alternatives? MLPs place activation functions on neurons, but can we instead place (learnable) activation functions on weights? Yes, we KAN! We propose Kolmogorov-Arnold Networks (KAN), which are more accurate and interpretable than MLPs.🧵

This is my favorite cell embedding paper! SATURN learns universal cell embeddings by coupling gene expression with pLM embeddings for cross-species integration of functionally-related genes (macrogenes). 🌟 Amazing for atlas integration! Paper: nature.com/articles/s4159… Code: github.com/snap-stanford/…

CRISPR-GPT: An LLM Agent for Automated Design of Gene-Editing Experiments Use Agent "to facilitate the process of selecting CRISPR systems, designing guide RNAs, recommending cellular delivery methods, drafting protocols, and designing validation experiments to confirm editing outcomes." biorxiv.org/content/10.110…

(1/5) Very excited to announce that our paper titled "Tandem mass spectrum prediction for small molecules using graph transformers" has been published in Nature Machine Intelligence! Paper: nature.com/articles/s4225… Code: github.com/Roestlab/massf… This method, MassFormer, was previously presented at the ISMB 2023 and Metabolomics 2023 conferences, and published as a preprint on Arxiv (arxiv.org/pdf/2111.04824…). It was led by my PhD student Adamo Young (@AdamoYoung), who is co-supervised by myself and Hannes Röst (@Roestlab) at the University of Toronto. 👇👇👇

if you are trying to get into deep learning for molecules like graph neural networks and equivariant neural networks check out dmol.pub/index.html if you haven't already @andrewwhite01 does an amazing job of building from the basics to E3 invariant models with e3nn (ch.19)

📢 New & improved material to dive into geometric deep learning! 💠🕸️ We (@mmbronstein @joanbruna @TacoCohen) delivered our Master's course on GDL @AIMS_Next once again & we make all materials publicly available! geometricdeeplearning.com/lectures/ See thread 🧵 for gems 💎 & dragons 🐉!

We’re proud to announce our method for drug response prediction, BiG DRP, has been published at Oxford Bioinformatics. Want to hear how we tackled the task? Read the thread 🧵👇 and the paper 📃 bit.ly/3n04UxV

A brief history of the Marcos family dynasty!

Ginusto n'yo yan. Gudlak na lang

Facebook should have a 'label' feature similar to how you label emails or cards in Trello, so that I could categorize my 'friends' according to when/where I met them and how stupid they are by voting a tax evader.

Our ICLR'22 paper (openreview.net/forum?id=YevsQ…) studies the phenomenon of dimensional collapsing in contrastive learning, in which the learned representation has a few dimensions collapsed to zero. Thanks all co-authors (@jingli9111, Pascal Vincent and @ylecun) for the great work! 1/

Here's an idea. Allow us to tap our debit/credit cards directly on the opus machine. Little things I miss about Korea

So last week I had to do an in-person tutorial but with zoom opened, because we were not given a room with a projector. This week they gave us a ancient projector that multiple laptops cannot seem to detect. 🙃🙃🙃🙃🙃🙃

According to my machine learning model, these 9 molecules are optimal. Please synthesize and test them. 😂

Large new cBioPortal data release: 3,680 samples from 6 published studies: pan-cancer data from #PCAWG (@Nature 2020), gliomas from the #GLASS Consortium (@Nature 2019), as well as several @sloan_kettering MSK-IMPACT studies. cbioportal.org/news

added to my reading list elifesciences.org/articles/60352

1/n I'm excited to share our preprint *Large-scale clinical interpretation of genetic variants using evolutionary data and deep learning* biorxiv.org/content/10.110… from a great collab with @NotinPascal @MafaldaFigDias @AidanNGomez @kpgbrock @yaringal (@OATML_Oxford) @deboramarks 🧵

New paper w/ @MarzyehGhassemi & @DrLukeOR where we argue against explainability for patient-level decision-making. tl;dr: Don’t ask for explainability at the bedside, it isn’t made for that. Paper: sciencedirect.com/science/articl…