Welcome to PyTrial documentation!

PyTrial is an easy-to-use Python package for a series of AI for drug development tasks. Clinical trial is the major step of the drug development process, where phase I, II, III, & IV trials are performed to comprehensively evaluate the efficacy and safety of a new drug.

PyTrial is featured for the following tasks and we are adding more!

  • Patient outcome prediction: predict the patient outcomes using tabular or sequential patient data.

  • Trial site selection: pick the best trial sites considering multiple objectives.

  • Trial outcome prediction: predict the trial outcomes using tabular or sequential trial data.

  • Patient-trial matching: match trials’ elibility criteria to patients’ EHRs for participant recruitment of clinical trials.

  • Trial similarity search: encode and retrieve similar clinical trials based on trial design documents.

  • Trial data simulation: generate synthetic EHRs or trial patient records in table or sequence.

_images/pytrial_tasks.png

The demonstration of clinical trial tasks supported by PyTrial.

If you find this package useful, please consider citing it in your scientific publications:

@misc{pytrial2022,
    title = {PyTrial: AI-driven In Silico Clinical Trial Optimization},
    author = {Wang, Zifeng and Theodorou, Brandon and Fu, Tianfan and Sun, Jimeng},
    year = {2022},
    month = {11},
    organization = {SunLab@UIUC},
    note = {Version 0.0.1},
}