/HER2_mutant

Primary LanguageJupyter Notebook

Live Cell Reporter Imaging

author: Nathaniel Evans
email: evansna@ohsu.edu

Sensitivity analysis use:

This will run many single-run analysis and save the results to disk.

$ ./HER2_sensitivity_runs.sh

NOTE: variables within HER2_sensitivity_runs.sh will need to be modified.

Single-run analysis use:

$ python HER2_classifier.py --data ./HER2_SKBR3_data_6-7-21/ --drug neratinib --sensitive_line WT --resistant_line T798I --load normalized --nclus 15 --out ./output/ --resample_sz 100 --burnin 0

--data

The outputs of Samuel's processing can be extracted to the necessary file structure using:

$ ./HER2_extract_data2.sh

NOTE: variables within HER2_extract_data2.sh will need to be changed for each run.

directory to data files, should be organized as:

/data_dir/ 
    /dataset_name/ 
        /normalized
            -> clover_all_cell.csv
            -> mscarlet_all_cell.csv
        /raw
            -> clover_all_cell.csv
            -> mscarlet_all_cell.csv

--drug

Can be trastuzumab or neratinib

--sensitive_line

The cell line to use as sensitive labels [WT]

--resistant_line

The cell line to use as resistant labels [T798I, ND611]

--load

Whether to use the normalized or raw data.

--nclus

The number of clusters to use.

--out

Directory path to save results to

--resample_sz

length of time series to resample to

--burnin

number of initial time points to ignore in analysis

Concordance Calls

Use HER2_sensitivity_results_analysis [<drug>].ipynb to aggregate the results of the sensitivity analysis, make concordance calls, and save results to file.

Conda environment

$ conda env create --file environment.yml