This repository is the official implementation of Panacea: A foundation model for clinical trial design, recruitment, search, and summarization. Model can be downloaded here.
See requirements.txt.
Here we reproduced all eight tasks across different settings in our code base, including trial design, patient-trial matching, trial search, and trial summarization.
We first use collected TrialAlign dataset to adapt Panacea to the vocabulary commonly used in clinical trials. Run the following
bash scripts/pretrain/run_pretrain_full.sh
Then, we conduct instruction-tuning step to enable Panacea to comprehend the user explanation of the task definition and the output requirement. Run
bash scripts/sft/sft.sh
Take patient-trial matching as an example, just run
bash scripts/eval/matching/patient2trial/panacea-7b.sh
To calculate the metrics, run
bash scripts/eval/matching/patient2trial/metrics/cls.sh
Evaluation of the other tasks is in the same way.
Please feel free to submit a Github issue if you have any questions or find any bugs. We do not guarantee any support, but will do our best if we can help.