server-scripts/qhtcp-workflow/apps/python/join_interactions.py

import pandas as pd
import os
import sys
import numpy as np

# Function to parse and set arguments
def parse_arguments():
    if len(sys.argv) == 1:  # Interactive mode
        args = [
            "/home/bryan/documents/develop/scripts/hartmanlab/workflow/out/20240116_jhartman2_DoxoHLD",
            2,
            "/home/bryan/documents/develop/scripts/hartmanlab/workflow/out/20240116_jhartman2_DoxoHLD/StudyInfo.csv",
            "/home/bryan/documents/develop/scripts/hartmanlab/workflow/out/20240116_jhartman2_DoxoHLD/Exp1",
            "/home/bryan/documents/develop/scripts/hartmanlab/workflow/out/20240116_jhartman2_DoxoHLD/Exp2"
        ]
    else:
        args = sys.argv[1:]

    return {
        "out_dir": os.path.abspath(args[0]),
        "sd": float(args[1]),
        "study_info": os.path.abspath(args[2]),
        "input_dirs": args[3:]
    }

args = parse_arguments()

# Create an array for the zscores files
def get_zscores_files(dirs):
    files = [os.path.join(study, "zscores", "zscores_interaction.csv") 
             for study in dirs if os.path.exists(os.path.join(study, "zscores", "zscores_interaction.csv"))]
    return files

zscores_files = get_zscores_files(args['input_dirs'])
print(f"The SD value is: {args['sd']}")

# Ensure there are enough files to compare
if len(zscores_files) < 2:
    sys.exit("Not enough experiments to compare, exiting script")

# Function to join zscores files
def join_zscores_files(files):
    joined_data = pd.read_csv(files[0])
    for file in files[1:]:
        temp_data = pd.read_csv(file)
        joined_data = pd.merge(joined_data, temp_data, on="OrfRep", how="outer")
    return joined_data

# Load and join zscores files
joined_data = join_zscores_files(zscores_files)

# Order and select columns
def order_and_select_columns(data):
    ordered_data = data[sorted(data.columns)]
    selected_headers = ordered_data.filter(regex="OrfRep|Gene|z_lm_k|z_shift_k|z_lm_l|z_shift_l")
    return selected_headers

selected_headers = order_and_select_columns(joined_data)

# Remove redundant columns like "Gene.1"
def clean_headers(data, suffixes):
    suffixes_to_remove = [f"Gene.{i}" for i in range(1, suffixes+1)]
    return data.drop(columns=suffixes_to_remove, errors='ignore')

headSel = clean_headers(selected_headers, len(zscores_files) - 1)
headSel2 = clean_headers(joined_data.filter(regex="OrfRep|Gene"), len(zscores_files) - 1)

# Fill NA values in Shift and Z_lm columns
def fill_na_in_columns(data):
    for column in data.columns:
        if "Shift" in column:
            data[column].fillna(0.001, inplace=True)
        elif "Z_lm_" in column:
            data[column].fillna(0.0001, inplace=True)
    return data

headSel = fill_na_in_columns(headSel)

# Filter based on standard deviation
def filter_by_sd(data, sd):
    if sd == 0:
        return data
    z_lm_cols = data.filter(regex="z_lm_")
    filter_vector = z_lm_cols.abs().ge(sd).any(axis=1)
    return data[filter_vector]

REMcRdy = filter_by_sd(headSel.filter(regex="OrfRep|Gene|z_lm_"), args['sd'])
shiftOnly = filter_by_sd(headSel.filter(regex="OrfRep|Gene|z_shift"), args['sd'])

# Reorder columns to interleave Z_lm and Shift data
def reorder_columns(data1, data2):
    combined_data = data1.copy()
    for i in range(2, data1.shape[1]):
        combined_data.insert(2 * i - 1, data2.columns[i], data2.iloc[:, i])
    return combined_data

combI = reorder_columns(headSel2, shiftOnly)

# Write output files
REMcRdy.to_csv(os.path.join(args['out_dir'], "REMcRdy_lm_only.csv"), index=False, quotechar=False)
shiftOnly.to_csv(os.path.join(args['out_dir'], "Shift_only.csv"), index=False, quotechar=False)

# Relabel headers using experiment names from StudyInfo.csv
def relabel_headers(headers, labels):
    new_labels = headers.copy()
    for i, header in enumerate(headers):
        suffix = header.split('.')[-1]
        if suffix.isdigit() and int(suffix) in range(1, 4):
            exp_name = labels.iloc[int(suffix) - 1, 1]
            new_labels[i] = header.replace(f".{suffix}", f"_{exp_name}")
    return new_labels

LabelStd = pd.read_csv(args['study_info'])

shiftOnly.columns = relabel_headers(shiftOnly.columns, LabelStd)
REMcRdy.columns = relabel_headers(REMcRdy.columns, LabelStd)

# Save relabeled files
REMcRdy.to_csv(os.path.join(args['out_dir'], "REMcRdy_lm_only.csv"), index=False, quotechar=False)
shiftOnly.to_csv(os.path.join(args['out_dir'], "Shift_only.csv"), index=False, quotechar=False)

# Save updated parameters
LabelStd.iloc[:, 3] = args['sd']
LabelStd.to_csv(os.path.join(args['out_dir'], "parameters.csv"), index=False)
LabelStd.to_csv(args['study_info'], index=False)