DeepTraj is a deep learning framework for modeling and predicting drug-induced gene expression dynamics over time and dose. By combining variational autoencoders with uncertainty aware B-splines, DeepTraj learns smooth, continuous gene-expression trajectories. This model was trained using the LINCS level 3 L1000 dataset (~1.8 million samples, "landmark" genes only).
Citation: pre-submission
Github: https://github.com/nathanieljevans/lincs-traj
Getting Started
Install environment
$ conda install environment.yaml
$ conda activate deep-traj
See tutorial.ipynb for usage details.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support