How to come up with research ideas? Excited about starting doing research but have no clue?🤷‍♂️🤷🏻‍♀️ Here are some simple methods that I found useful in identifying initial directions. Check out the thread below 👇

Google announces Dreamix: a model that generates videos when given: - video + prompt (Video editing) - input images + prompt (Subject Driven Generation) - input image + prompt (Image-toVideo

Want to understand what Spatial Transcriptomics (ST) is,but unsure where to start? Need to analyze spatial genomics data,but unclear how? Here's how to start👇 An introductory #Bioinformatics review on: 1. ST Technologies 2. Exploratory data analysis 3. Biological implications

🤞ColabFold v1.5.2🤞 This update attempts to fix a system memory leak when running predictions on large proteins/complexes. Thx to liyv (from discord) to helping debug! (1/3)

Roughly one year ago we started out venturing in the world of #proteindesign. Today I am very happy to share our first preprint ‘Efficient and scaleable de-novo protein design using a relaxed sequence space’ (biorxiv.org/content/10.110…) with Sergey Ovchinnikov & @hendrik_dietz 1/x

AlphaFold inverted to hallucinate denovo proteins of up to 600 amino acids in length🤯 (animation below shows the designed protein docked into CryoEM density) Exciting work with: Christopher Frank, Ali Khoshouei, Yosta de Stigter, Dominik Schiewitz, Shihao, @hendrik_dietz

Puzzle: The residue index encodes the position embedding for models like AlphaFold. This residue index is converted into an offset matrix. (1/3)

Our new work with Sergey Ovchinnikov Institute for Protein Design core labs, DiMaio lab. We introduce AfCycDesign, where we adapt the #AlphaFold network for cyclic peptide structure prediction and design. biorxiv.org/content/10.110…

There are other protein designers, but I think Sergey Ovchinnikov is the only protein magician. This is the latest in a series of genius hacks (and I mean this in the best possible spirit of the term) from him and his collaborators, in this case Gaurav Bhardwaj & team at Institute for Protein Design.

Efficiently generate de novo proteins by - optimizing residue logits for max AF confidence - redesigning the sequence using ProteinMPNN Tested in the lab, including CryoEM structures Christopher Frank Ali Khoshouei Sergey Ovchinnikov @hendrik_dietz biorxiv.org/content/10.110…

🚨Stop using positional encoding (PE) in Transformer decoders (e.g. GPTs). Our work shows 𝗡𝗼𝗣𝗘 (no positional encoding) outperforms all variants like absolute, relative, ALiBi, Rotary. A decoder can learn PE in its representation (see proof). Time for 𝗡𝗼𝗣𝗘 𝗟𝗟𝗠𝘀🧵[1/n]

ColabDock: inverting AlphaFold structure prediction model for protein-protein docking with experimental restraints biorxiv.org/cgi/content/sh… #biorxiv_bioinfo

"ColabDock: inverting AlphaFold structure prediction model for protein-protein docking with experimental restraints" Use experimental restraints as losses & back-propagating to input sequences. Wonder how many sequence changes are introduced to fit data? biorxiv.org/content/10.110…

I'm excited to share that I'll be joining MIT Biology as an Asst Prof. in Jan 2024! Come join us! 🤓🧪🖥️🧬

Now out in slightly shorter form :) nature.com/articles/s4158… Kotaro Tsuboyama Justas Dauparas Elodie Laine Sergey Ovchinnikov