Computational workflow for design of dose-response experiments
- The repository can be installed from command line as shown below
$ git clone https://github.com/datarail/datarail.git
- To install dependencies and enable importing modules from any location on your local machine,
datarailfolder, followed by the command below.
$ pip install -e .
- Set up the well and plate level metadata files as shown in
- Start a Jupyter notebook or IPython session.
- The layout of drugs on doses across 96/384 well plates can be constructed using the code below. The pandas dataframe
dfmcontains the desingned layout. Refer to
datarail/examplesfor a detailed explanation with examples.
import pandas as pd from datarail.experimental_design import process_assay as pa dfp = pd.read_csv('plate_level_metadata.csv') dfm = pa.randomize_wells(dfp) dfm.to_csv('dose_response_layout_metadata.csv', index=False)
- The metadata file can be exported to a .hpdd file that can be used by the D300 printer. The stock concentraion for the drug also needs to be provided for each drug in the assay.
from datarail.experimental_design import hpdd_utils as hu hu.export_hpdd(dfm, dfs, 'dose_response_layout_metadata.hpdd')
- The layout can be visualized using the code below
from datarail.experimental_design import plot_plate_layout as ppl ppl.plot_summary(dfr, 'dose_response_layout_metadata.pdf')
- If the D300 software was used for designing the experiment, follow the steps below inorder to save the metadata file based on
datarailconvention for subsequent downstream analysis.
- Open the
.xmlfrom D300 in Excel and save as a
- Use the code below to save the metadata in a dataframe
from datarail.experimental_design import export_D300_xml as edx dfm = edx.export2pd('D300_filename.xlsx')
- Open the