call for papers

 We have two tracks: application track and ML track 

Application Track

AI for DNA, RNA, and cell and gene therapeutics, which leverages cutting-edge AI methods. For example, 

• AI for therapeutic RNAs (e.g., optimizing UTR/codons to enhance translational efficiency for mRNA vaccines, designing antisense oligonucleotides by optimizing stability).

• AI for cell and gene therapies (e.g., designing tissue/cell-type-specific regulatory elements, selecting suitable cells for therapy, AI-based CRISPR design) 

• AI for protein engineering

• New representation of molecules

• Designing delivery systems (e.g., design of nanoparticles for efficient delivery of RNA/DNA therapeutics to target cells and tissues).

• Peptide therapeutics, microbiome-based therapies 

• Usage of FMs for agents in drug discovery 

 We also welcome papers that focus on established modalities, such as antibodies and small molecules, especially with new AI techniques.

ML track (Foundational Models for Drug Discovery) 

Foundational models (FMs) typically refer to large-scale predictive or generative models trained on extensive datasets. Our goal is to explore the importance of FMs in drug development, i.e., design and target identification. For example, 

• Method that bridges the gap between FMs and design in drug discovery/target identification. 

• New AI-driven design approaches (e.g., utilizing various representations such as primary/tertiary structures of RNAs, employing state-of-the-art RL, miniaturizing DNAs). 

• Large-scale predictive/generative models for new modalities (e.g., (1) leveraging biological knowledge, (2) modeling multimodal aspects such as DNA/RNA/protein or 2D/3D structures, (3) new diffusion models (DMs)/long-range neural network architecture models for biological sequences) 

• Fine-tuning foundational models from lab feedback

• Interpretability in foundational models (e.g., knowledge graphs, retrieval augmented generation). 

• Foundation models employing multi-modal perturbation (genetic/molecular perturbation), multimodal readouts  (transcriptomic, phenotypic readouts), and multi-parameter assays.

Dates 

The submission deadline is October 1, 2024 (Anywhere on Earth). Acceptance notifications will be sent by October 9, 2024.

Submission Instructions

• Submissions must be made via OpenReview and must be anonymized. 

• Full submissions may be up to 9 pages long (excluding references and appendices) and should use the NeurIPS/ICLR template. 

• Submissions should present original research (i.e., not accepted in other conferences or journals). Authors can upload concise versions of parallel submissions to other conferences, such as NeurIPS or ICLR. Dual submissions are accepted, but we discourage submitting to multiple NeurIPS workshops. 

• All accepted papers will be available on the workshop website, but If desired, you may choose to keep your paper private. No formal workshop proceedings will be published.

For any questions, email us at aidrugx@googlegroups.com 

For Accepted Papers (ReVISION) 

• The revision deadline is 11/20, 2024. No Page Limit. 

• Please use the template in https://www.overleaf.com/read/nwfndtjwcdbx#18cacd

• Technically, our paper is non-archival, although accepted papers will be made publicly accessible by default. If you prefer not to make it public, we recommend uploading a version containing only the abstract.