glm-pca, normalization factor results

norm_results.py

import scanpy as sc, numpy as np
from matplotlib import pyplot as plt
from sklearn.utils import sparsefuncs

def norm_frac(adata):
    sparsefuncs.inplace_row_scale(adata.X, 1 / adata.X.sum(axis=1).A1)
    return adata

def norm_cpm(adata):
    sparsefuncs.inplace_row_scale(adata.X, 1_000_000 / adata.X.sum(axis=1).A1)
    return adata

def norm_median(adata):
    sc.pp.normalize_total(adata)
    return adata

!wget https://cf.10xgenomics.com/samples/cell-exp/1.1.0/cd14_monocytes/cd14_monocytes_filtered_gene_bc_matrices.tar.gz
!tar xzf cd14_monocytes_filtered_gene_bc_matrices.tar.gz

mono = sc.read_10x_mtx("./filtered_matrices_mex/hg19/")

sc.pp.filter_genes(mono, min_cells=1)
sc.pp.calculate_qc_metrics(mono, inplace=True)

adatas = {
    "cpm": norm_cpm(mono.copy()),
    "frac": norm_frac(mono.copy()),
    "median": norm_median(mono.copy()),
}

for k, adata in adatas.items():
    sc.pp.log1p(adata)
    sc.pp.pca(adata)
    sc.tl.tsne(adata)

with plt.rc_context({"figure.dpi": 120, "figure.figsize": (18, 4)}):
    fig, axs = plt.subplots(ncols=3)
    for (k, adata), ax in zip(adatas.items(), axs):
        sc.pl.tsne(
            adata, 
            color="zero_frac", 
            title=f"{k}, tsne", 
            ax=ax,
            show=False, 
            vmin=0.97, 
            size=30, 
            cmap="RdBu_r"
        )