Comparison of Data Distribution Before and After Adding Noise

Loading Packages

import warnings
## filter out all warnings
warnings.filterwarnings('ignore')

import scanpy as sc
import numpy as np
import matplotlib.pyplot as plt

Loading Data

Combination1_id = 4     ## 1, 2, 3, 4

adata_RNA = sc.read_h5ad(f'../../../Data/Noise_Combination_{Combination1_id}/Combination{Combination1_id}_RNA.h5ad')
adata_ADT = sc.read_h5ad(f'../../../Data/Noise_Combination_{Combination1_id}/Combination{Combination1_id}_Protein.h5ad')

Ploting

RNA_DATA = [adata_RNA.X, adata_RNA.obsm['level_1'], adata_RNA.obsm['level_2'], adata_RNA.obsm['level_3']]
ADT_DATA = [adata_ADT.X, adata_ADT.obsm['level_1'], adata_ADT.obsm['level_2'], adata_ADT.obsm['level_3']]

fontsize = 30

for level_id in range(4):
    RNA = RNA_DATA[level_id]
    protein = ADT_DATA[level_id]
    RNA = np.average(RNA, axis=1)
    protein = np.average(protein, axis=1)

    plt.figure(figsize=(7, 14))

    # RNA
    ax1 = plt.subplot(2, 1, 1)
    hist = plt.hist(RNA, bins=20, color='#fc9e4f', alpha=1)
    plt.title('')
    plt.xlabel('Average RNA expression', fontsize=fontsize, fontname='Arial')
    plt.ylabel('Frequency', fontsize=fontsize, fontname='Arial')
    plt.tick_params(axis='both', labelsize=fontsize)
    ticks = np.arange(-0.5, 1.31, 0.6)
    plt.xticks(ticks)
    ax1.grid(True)
    ax1.spines['top'].set_linewidth(1.5)
    ax1.spines['right'].set_linewidth(1.5)
    ax1.spines['bottom'].set_linewidth(1.5)
    ax1.spines['left'].set_linewidth(1.5)

    # Protein
    ax2 = plt.subplot(2, 1, 2)
    plt.hist(protein, bins=40, color='#07beb8', alpha=1)
    plt.title('')
    plt.xlabel('Average protein expression', fontsize=fontsize, fontname='Arial')
    plt.ylabel('Frequency', fontsize=fontsize, fontname='Arial')
    ticks = np.arange(-1, 7.1, 2)
    plt.xticks(ticks)
    plt.tick_params(axis='both', labelsize=fontsize)
    ax2.grid(True)
    ax2.spines['top'].set_linewidth(1.5)
    ax2.spines['right'].set_linewidth(1.5)
    ax2.spines['bottom'].set_linewidth(1.5)
    ax2.spines['left'].set_linewidth(1.5)

    plt.tight_layout()
    plt.subplots_adjust(hspace=0.3)

    plt.savefig(f'Combination{Combination1_id}/Noise_Level_{level_id}.png', dpi=500)
    plt.savefig(f'Combination{Combination1_id}/Noise_Level_{level_id}.eps')

    plt.show()

Saving the Results

import pandas as pd
import numpy as np
import scanpy as sc

with pd.ExcelWriter('Histogram_Frequency_Data.xlsx', engine='openpyxl') as writer:
    for combo_id in range(1, 5):  # Combination 1-4
        # Loading data
        adata_RNA = sc.read_h5ad(f'../../../Data/Noise_Combination_{combo_id}/Combination{combo_id}_RNA.h5ad')
        adata_ADT = sc.read_h5ad(f'../../../Data/Noise_Combination_{combo_id}/Combination{combo_id}_Protein.h5ad')

        RNA_DATA = [adata_RNA.X, adata_RNA.obsm['level_1'], adata_RNA.obsm['level_2'], adata_RNA.obsm['level_3']]
        ADT_DATA = [adata_ADT.X, adata_ADT.obsm['level_1'], adata_ADT.obsm['level_2'], adata_ADT.obsm['level_3']]

        combo_data = []

        for level_id in range(4):
            RNA = np.average(RNA_DATA[level_id], axis=1)
            protein = np.average(ADT_DATA[level_id], axis=1)

            rna_counts, rna_bins = np.histogram(RNA, bins=20)
            protein_counts, protein_bins = np.histogram(protein, bins=40)

            for i in range(len(rna_counts)):
                combo_data.append({
                    'Noise_Level': level_id,
                    'DataType': 'RNA',
                    'Bin_Index': i + 1,
                    'Bin_Start': rna_bins[i],
                    'Bin_End': rna_bins[i+1],
                    'Bin_Center': (rna_bins[i] + rna_bins[i+1]) / 2,
                    'Frequency': rna_counts[i],
                    'Bin_Width': rna_bins[i+1] - rna_bins[i]
                })

            for i in range(len(protein_counts)):
                combo_data.append({
                    'Noise_Level': level_id,
                    'DataType': 'Protein',
                    'Bin_Index': i + 1,
                    'Bin_Start': protein_bins[i],
                    'Bin_End': protein_bins[i+1],
                    'Bin_Center': (protein_bins[i] + protein_bins[i+1]) / 2,
                    'Frequency': protein_counts[i],
                    'Bin_Width': protein_bins[i+1] - protein_bins[i]
                })

        # Save
        df = pd.DataFrame(combo_data)
        df.to_excel(writer, sheet_name=f'Combination_{combo_id}', index=False)