I am so thrilled to announce that I am starting as an Assistant Professor this fall at UMass Amherst CICS @manningcics. Our research group will build and apply computational models for bacterial genomics, focused on antibiotic resistance and machine learning. Recruiting soon!
Can protein LMs reveal scientific knowledge? We start by asking how pLMs turn sequences into structure signals. We map a contact prediction circuit: early motif features gate later domain features. Spurious or science? We can now test. 🧵(1 of N)
@manningcics at @UMassAmherst is planning to hire two tenure-track faculty in #NLProc at Assistant / Associate Professor levels.
Job posting: https://t.co/ZTImEqtEtO
Applications are due on Dec. 1st.
If you have any questions, please feel free to reach out!
Please RT!
If you're interested in a PhD applying stat/AI methods to prediction of phage susceptibility & cocktail design with me, @edzewestra + @StinekevanHoute , we're advertising an MRC funded project - deadline 4th November. International students welcome!
https://t.co/4sQabeK7va
The 2025 Hertz Fellowship application is open! In addition to 5 years of funding, Hertz Fellows join a community of peers, including some of the nation’s most noted #science & #technology leaders. Start your application today! https://t.co/YBbF6acnFF
Super happy that our recent paper on bloodstream Klebsiella is featured in @NatureComms Editors' Highlights in Microbiology and Infectious Diseases https://t.co/RWlrphRl1W
Introducing the 😱OMG dataset and 🤖gLM2! OMG is a massive Open MetaGenomic corpus totaling 3.1Tbp of data and 3.3B proteins. We train gLM2, the first mixed-modality genomic language model, on OMG. 1/🧵 #BioML#AI4Science
Does composting spread antibiotic resistance? Researchers found that the release and transmission of extracellular antibiotic resistant genes (ARGs) during cow manure composting is facilitated by heat and the lysis of bacteria by phages. https://t.co/QTASKf7At8 @AppEnvMicro
The MLSB 2024 call for papers is live at https://t.co/wv9C4we1CQ !
New this year: we're excited to use MLSB to highlight importance of benchmarks. We'll have 2 tracks for models evaluated on https://t.co/X34vSEgXA4 / https://t.co/0tTgKxEK4m from @vant_ai @nvidia@MIT_CSAIL !
Great review about the "AlphaFold Mania" 🔥:
- structure prediction
- remote homology relationships
- molecular replacement
- Interactomes
- variant effect predictor
- conformational sampling
- and much more
AlphaFold two years on - validation and impact
https://t.co/7WWIBVU9O3
Accurate prediction of protein function using statistics-informed graph networks | @NatureComms
- PhiGnet, uses evolutionary couplings (EVCs) and residue communities (RCs) with Dual-Channel Graph Convolutional Networks. Embed sequences via ESM-1b
- Use Grad-CAM method computes activation scores for functional residues Interpretability
Link: https://t.co/iUHrqbtV0j
Which Strains of Tuberculosis Are the Most Infectious?
New research from DBMI's @MahaFarhat and @iMGroschel in @NatureMicrobiol shows shared geographic origin between TB strain and human host could amplify risk for infection. Story via @harvardmed. https://t.co/g85G5iw6Ys
It's great to see more work being done on *contextualizing* protein embeddings with genomic context information! Check out this new viral gLM work by @pCodyM et al!🎉