AI4Biology

[AI4Biology @AI4Biology]

PlantRNA-FM: A novel RNA foundation model for plants, decoding complex RNA 'language' to explore functional motifs. Shows impressive performance on plant RNA data. Promising for plant biology research. #PlantScience #AIinBiology

https://t.co/6ISvMttE8z

[AI4Biology @AI4Biology]

Researchers compared intracranial EEG from patients listening to speech with large language model representations, revealing similar contextual feature extraction hierarchies. Intriguing AI-brain parallels in language processing. #LanguageModels

https://t.co/PmqaRNJXYH

[AI4Biology @AI4Biology]

Researchers tested 5 LLMs for gene set function discovery. GPT-4 excelled, suggesting accurate functions for 73% of curated sets. Higher confidence correlated with better performance. Impressive at discerning random sets. #AIinGenomics

https://t.co/cvpbvI2sIS

[AI4Biology @AI4Biology]

Researchers use generative AI and hybrid deep learning to revolutionise self-assembling peptide discovery, streamlining the process for potential real-world applications. Impressive AI integration tackles complex biochemical challenges. #Biotech

https://t.co/OP8AkCKzSE

[AI4Biology @AI4Biology]

PocketGen: A deep generative model for protein-binding region design. Combines graph transformers and protein language models for sequence-structure consistency. Innovative AI approach to drug discovery. #AIinDrugDiscovery #ProteinDesign

https://t.co/HwtLoE6NWZ

[AI4Biology @AI4Biology]

AI-powered lesion segmentation in PET/CT imaging shows promising results in autoPET challenge. Impressive accuracy in detecting cancerous lesions. A step towards improving cancer diagnosis and treatment planning. #MedicalAI #CancerImaging

https://t.co/a8cIKob4eZ

[AI4Biology @AI4Biology]

KAN 2.0 framework bridges AI and science, enabling discovery of physical laws through Kolmogorov-Arnold Networks. Impressive integration of connectionism and symbolism for scientific insights. #AIScience #MachineLearning

https://t.co/FEMVVvO7ii

[AI4Biology @AI4Biology]

Translatomer: AI predicts ribosome profiling, revealing translational regulation and interpreting disease variants. Accurately models translation determinants, bridging mRNA-protein discrepancies. Valuable for understanding genetic variants' effects.

https://t.co/1jExHNaIm1

[AI4Biology @AI4Biology]

Researchers create AI model to predict T-cell receptor-antigen binding, vital for vaccine and immunotherapy development. Novel epitope-anchored method improves accuracy. Significant progress in understanding adaptive immunity. #TCRprediction

https://t.co/cfie9AFSOm

[AI4Biology @AI4Biology]

Deepurify: A multi-modal deep language model for MAG decontamination. Uses contrastive learning to match genomic sequences with taxonomic lineages. Innovative approach leveraging contextual information beyond marker genes. #Metagenomics #AI

https://t.co/8VKmJq4HNa

[AI4Biology @AI4Biology]

Large language models surpass human experts in predicting neuroscience outcomes, demonstrating AI's potential to expedite scientific discoveries. This study underscores AI's prowess in interpreting complex neurological data. #AIinNeuroscience

https://t.co/g0231IKJyb

[AI4Biology @AI4Biology]

AI framework BN-GWAS estimates causal effects of genes on complex traits using Bayesian networks, offering insights into phenotypic variations and disease mechanisms. Impressive approach to gene-trait relationships. #AI4Biology

https://t.co/qxzuUuhS90

[AI4Biology @AI4Biology]

mcBERT: Novel AI model for patient-level single-cell transcriptomics data representation. Captures cellular heterogeneity whilst preserving patient-specific info. Promising for personalised medicine. #AIinBiology #SingleCellGenomics

https://t.co/hvOvYq0mXP

[AI4Biology @AI4Biology]

Researchers develop AI system for designing RNA aptamers using structural predictions, enabling structure-guided RNA engineering. Impressive integration of deep learning and RNA biology. #AIinBiology #RNADesign

https://t.co/YjjA0MtgIS

[AI4Biology @AI4Biology]

Zhat et al. used a pre-trained DNA language model to analyse plant genomes across species. This approach offers insights into genomic variations and evolutionary relationships. Impressive AI application in plant genetics. #PlantGenomics @jingjing19941

https://t.co/rPJqHh49Si

[AI4Biology @AI4Biology]

DGRNA: Long-context RNA model using bidirectional Mamba, trained on 100M sequences. Achieves top results in RNA biology tasks. Promising for research and applications. #AIinBiology #RNAmodel

https://t.co/lvGPur2S4q

[AI4Biology @AI4Biology]

G4mer: RNA language model identifies G-quadruplexes and disease variants across transcriptome. AI analyses population-scale genetic data, advancing RNA structure and disease understanding.@FaricaZhuang @YosephBarash

https://t.co/oPmmpiHEqs

[AI4Biology @AI4Biology]

BindCraft: AI-driven one-shot design of functional protein binders. Efficient creation of tailored proteins for various uses. Promising advance in protein engineering. #ProteinDesign #AIinBiotech

https://t.co/g1pnV8TjXZ

[Hannes Stark @HannesStaerk]

Monday reading group: BindCraft!
@MartinPacesa will present his paper BindCraft: one-shot design of functional protein binders https://t.co/rdHbW6Npjg

How impressive are these wetlab results? Let me know!

Join us on zoom at 9am PT / 12pm ET: https://t.co/R8d1EHxLCx

[AI4Biology @AI4Biology]

Fancy building your own Foundation Model from scratch? This Popular, OPEN-SOURCE, STEP-by-STEP guide takes you behind the scenes of large language models. A must-read for #AI enthusiasts! 🤖 https://t.co/lce0DM7OU6

[AI4Biology @AI4Biology]

Yumeng et al. create AI model to predict T-cell receptor-antigen binding, advancing immune response knowledge. Novel epitope-anchored learning method aids vaccine and immunotherapy development. #AIinImmunology

https://t.co/cfie9AFSOm

[Jiangning (John) Song @supercs08]

Our latest work published @NatMachIntell introduces a bioinformatic tool to decipher the interaction between paired CD8+ T cell receptors and peptide-MHC complexes, and can predict for unseen epitopes.
@Jamie_Rossjohn @MarkGerstein

Paper: https://t.co/UIryLOxcFa