01 - NGS collection [mouse]

Timepoints were adjusted based on Saiz and Plusa et al., Reproduction 2013

Mouse: - Guo et al., 2010: qPCR on steroids, complicated, skipping - maybe for prediction, could be interesting) - Guo et al., 2014: Bisulfite-Seq (skipping) - ~~Ohnishi et al., 2014: Microarray (used in ccmnetplus for validation) *~~

!which pip

/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/bin/pip

import anndata
import pandas as pd
import scanpy as sc
import pandas as pd
import seaborn as sns


def load_experiment(filename: str, GEO: str, left_on: str = 'sample', right_on='SRX'):
    adata = sc.read(filename)
    metadata = pd.read_csv(f"../pipeline/fetchngs/{GEO}_metadata.csv")

    # sometimes we have duplicates because the ID is not unique
    adata.obs = pd.merge(adata.obs, metadata, left_on=left_on, right_on=right_on)\
                    .drop_duplicates(left_on)\
                    .set_index(left_on)

    adata.obs['ct_orig'] = adata.obs['sample_title']

    return adata

%run ../scripts/helpers.py

/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/scvi/_settings.py:63: UserWarning: Since v1.0.0, scvi-tools no longer uses a random seed by default. Run `scvi.settings.seed = 0` to reproduce results from previous versions.
  self.seed = seed
/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/scvi/_settings.py:70: UserWarning: Setting `dl_pin_memory_gpu_training` is deprecated in v1.0 and will be removed in v1.1. Please pass in `pin_memory` to the data loaders instead.
  self.dl_pin_memory_gpu_training = (
/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
  from .autonotebook import tqdm as notebook_tqdm

def normalize_smartseq(adata: sc.AnnData, gene_len: pd.DataFrame) -> sc.AnnData:
    print("SMART-SEQ: Normalization")

    common_genes = adata.var_names.intersection(gene_len.index)
    print(f"SMART-SEQ: Common genes {common_genes.shape[0]}")

    lengths = gene_len.loc[common_genes, "length"].values
    normalized = adata[:, common_genes].copy()
    normalized.X = normalized.X / lengths * np.median(lengths)
    normalized.X = np.rint(normalized.X)

    return normalized

# pip install gtftools
# !gtftools \
#     -l ../data/external/Mus_musculus_GRCm38_102_gene_length.txt \
#     /scratch/Brickman/references/mus_musculus/ensembl/GRCm38_102/Mus_musculus.GRCm38.102.gtf

genes_length = pd.read_table("../data/external/Mus_musculus_GRCm38_102_gene_length.txt").set_index('gene')
genes_length['length'] = genes_length['mean']
genes_length

	mean	median	longest_isoform	merged	length
gene
ENSMUSG00000102693	1070	1070	1070	1070	1070
ENSMUSG00000064842	110	110	110	110	110
ENSMUSG00000051951	3592	3634	4153	6094	3592
ENSMUSG00000102851	480	480	480	480	480
ENSMUSG00000103377	2819	2819	2819	2819	2819
...	...	...	...	...	...
ENSMUSG00000117740	331	331	331	331	331
ENSMUSG00000117782	1570	1570	1570	1570	1570
ENSMUSG00000117951	256	256	256	256	256
ENSMUSG00000095993	168	168	168	168	168
ENSMUSG00000118197	243	243	243	243	243

55364 rows × 5 columns

1 Datasets

Dataset	Technology	Download	Notes
Deng et al., 2014	SMART-seq 1/2	GSE45719	Done
Biase et al., 2014	SMART-seq	GSE57249	Done
Posfai et al., 2017	SMART-seq	GSE84892	Done
Goolamn et al., 2016	SMART-seq2	E-MTAB-3321	Done
Boroviak et al., 2015	SMART-seq2	Sup Table S1	Done
Chen et al., 2016	SMART-seq2	GSE74155	Done
Nowotschin et al., 2019	10X v2	Portal	Done
Fan et al., 2015	SUPeR-seq	GSE53386	Done
Borensztein et al., 2017	qRT-PCR	GSE80810	Done
Stirparo et al., 2020	SMART-seq2	GSE159030	Done
Xue et al., 2013	qRT-PCR	GSE44183	Done
Yanagida et al., 2022	SMART-seq2	GSE148462	Done
Mohammed et al., 2017	SMART-seq2	GSE100597	Done

1.1 Deng et al., 2014 GSE45719

nf-core_tower.sh \
    Deng_et_al_2014 \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE45719.txt

nf-core_tower.sh Deng_et_al_ALIGNED nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --protocol smartseq \
    --aligner star \
    --fasta /scratch/Brickman/references/mus_musculus/ensembl/GRCm38_102/Mus_musculus.GRCm38.dna_sm.primary_assembly.fa \
    --gtf /scratch/Brickman/references/mus_musculus/ensembl/GRCm38_102/Mus_musculus.GRCm38.102.gtf \
    --star_index /scratch/Brickman/references/mus_musculus/ensembl/GRCm38_102/star_2.7.10a \
    --input /scratch/Brickman/pipelines/Deng_et_al_2014/results/samplesheet/samplesheet.csv

1.1.1 Preparation

deng_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE45nnn/GSE45719/matrix/GSE45719_series_matrix.txt.gz", 
                              skiprows=30, index_col=0).T

deng_metadata['ct'] = deng_metadata['!Sample_source_name_ch1'].values
deng_metadata['SRX'] = deng_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')
deng_metadata = deng_metadata.reset_index()

legend = {
    '16-cell stage blastomere': '16C', 
    '2-cell stage blastomere': '2C',
    '4-cell stage blastomere': '4C',
    '8-cell stage blastomere': '8C',
    'Early 2-cell stage blastomere (31-32h post-fertilization)': '2C',
    'Early blastocyst cell (86-88h post-fertilization)': 'Blastocyst',
    'Late 2-cell stage blastomere (46-48h post-fertilization)': '2C',
    'Late blastocyst cell (100-102h post-fertilization)': 'Blastocyst',
    'Liver': 'Liver',
    'Liver cell': 'Liver', 
    'Mid 2-cell stage blastomere (34-40h post-fertilization)': '2C',
    'Mid blastocyst cell (92-94h post-fertilization)': 'Blastocyst',
    'Zygote': 'Zygote',
    'fibroblast (primary culture from tail)': 'Fibroblast'
}

for old_name, new_name in legend.items():
    deng_metadata.loc[deng_metadata.ct == old_name, 'ct'] = new_name
deng_metadata.ct = deng_metadata.ct.astype('category')
deng_metadata = deng_metadata.query('ct not in ["Liver", "Blastocyst", "Fibroblast"]').copy()

deng_metadata['index'].to_csv("../pipeline/fetchngs/GSE45719.txt", index=None, header=None)
deng_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE45719_metadata.csv")

1.1.2 Load dataset

deng = load_experiment(
    "../data/external/aligned/mouse/Deng_et_al_ALIGNED/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE45719"
)

deng.obs['experiment'] = "Deng et al., 2014"
deng.obs['technology'] = 'SMART-seq'
deng.obs.loc[deng.obs.sample_alias.isin([f'GSM12780{i:02d}' for i in range(9, 45)]), 'technology'] = 'SMART-seq2'
deng.obs['batch'] = "DENG_1"
deng.obs.loc[deng.obs.technology == 'SMART-seq2', 'batch'] = "DENG_2"
deng.obs = deng.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

deng = normalize_smartseq(deng, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

1.2 Biase et al., 2014 GSE57249

nf-core_tower.sh \
    Biase_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE57249.txt

nf-core_tower.sh Biase_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Biase_et_al/results/samplesheet/samplesheet.csv

1.2.1 Preparation

biase_metadata = pd.concat([
    pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE57nnn/GSE57249/matrix/GSE57249-GPL13112_series_matrix.txt.gz",
                  skiprows=33, index_col=0).T.reset_index().set_index('!Sample_geo_accession'),
    pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE57nnn/GSE57249/matrix/GSE57249-GPL17021_series_matrix.txt.gz",
                  skiprows=33, index_col=0).T.reset_index().set_index('!Sample_geo_accession')
])

biase_metadata['ct'] = biase_metadata['index'].values
biase_metadata['SRX'] = biase_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')
biase_metadata = biase_metadata.reset_index()

legend = {
    'Zygote': 'Zygote',
    'Two-cell': '2C',
    'Four-cell': '4C',
    'Inner cell mass': 'E3.5-ICM',
    'Trophectoderm': 'E3.5-TE'
}

for old_name, new_name in legend.items():
    biase_metadata.loc[biase_metadata.ct.str.startswith(old_name), 'ct'] = new_name
biase_metadata.ct = biase_metadata.ct.astype('category')

biase_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE57249.txt", index=None, header=None)
biase_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE57249_metadata.csv")

1.2.2 Load dataset

biase = load_experiment(
    "../data/external/aligned/mouse/Biase_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE57249")

biase.obs['batch'] = "BIASE_1"
biase.obs['experiment'] = "Biase et al., 2014"
biase.obs['technology'] = "SMART-seq2"

biase.obs = biase.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

biase = normalize_smartseq(biase, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

biase

AnnData object with n_obs × n_vars = 56 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.3 Posfai et al., 2017 GSE84892

nf-core_tower.sh \
    Posfai_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE84892.txt

nf-core_tower.sh Posfai_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Posfai_et_al/results/samplesheet/samplesheet.csv

1.3.1 Preparation

posfai_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE84nnn/GSE84892/matrix/GSE84892_series_matrix.txt.gz",
                                skiprows=31, index_col=0).T

posfai_metadata['SRX'] = posfai_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')
posfai_metadata = posfai_metadata.reset_index()

legend = {
    '16': '16C-',
    '32': 'E3.25-',
    '64': 'E3.5-'
}

posfai_metadata['ct'] = 'UNKNOWN'
for old_name, new_name in legend.items():
    posfai_metadata.loc[posfai_metadata.loc[:, '!Sample_characteristics_ch1'].iloc[:, 0].str.contains(old_name), 'ct'] = new_name

posfai_metadata.ct = posfai_metadata.ct + posfai_metadata.loc[:, '!Sample_characteristics_ch1'].iloc[:, 2].str.replace('lineage: ', '')
posfai_metadata = posfai_metadata.query('~ct.str.contains("CO") or ct != "UNKNOWN"').copy()
posfai_metadata.loc[posfai_metadata.ct.str.contains('16C'), 'ct'] = '16C'

posfai_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE84892.txt", index=None, header=None)
posfai_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE84892_metadata.csv")

1.3.2 Load dataset

posfai = sc.read("../data/external/aligned/mouse/Posfai_et_al/results/star/mtx_conversions/combined_matrix.h5ad")
posfai = posfai[~posfai.obs.duplicated()].copy()
posfai_metadata = pd.read_csv("../pipeline/fetchngs/GSE84892_metadata.csv")

posfai.obs = pd.merge(posfai.obs, posfai_metadata, left_on="sample_title", right_on="index").set_index('sample_title')

posfai = posfai[posfai.obs.ct != "E3.25-CO"].copy()

/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/anndata/_core/anndata.py:1830: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.
  utils.warn_names_duplicates("obs")

posfai.obs['batch'] = "POSFAI_1"
posfai.obs['experiment'] = "Posfai et al., 2017"
posfai.obs['technology'] = "SMART-seq2"
posfai.obs['ct_orig'] = posfai.obs['!Sample_characteristics_ch1']

posfai.obs = posfai.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

posfai = normalize_smartseq(posfai, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

posfai

AnnData object with n_obs × n_vars = 261 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.4 Goolamn et al., 2016 E-MTAB-3321

nf-core_tower.sh \
    Goolam_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/E-MTAB-3321.txt

nf-core_tower.sh Goolam_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Goolam_et_al/results/samplesheet/samplesheet.csv

1.4.1 Preparation

goolamn_metadata = pd.read_table("https://www.ebi.ac.uk/biostudies/files/E-MTAB-3321/E-MTAB-3321.sdrf.txt")

goolamn_metadata['ct'] = goolamn_metadata['Characteristics[developmental stage]'].values

legend = {
    'cleavage 2-cell' : '2C',
    'cleavage 4-cell': '4C',
    'cleavage 8-cell': '8C',
    'cleavage 16-cell': '16C',
    'cleavage 32-cell': '32C'
}

for old_name, new_name in legend.items():
    goolamn_metadata.loc[goolamn_metadata.ct == old_name, 'ct'] = new_name
goolamn_metadata.ct = goolamn_metadata.ct.astype('category')

goolamn_metadata = goolamn_metadata.query('ct != "32C"').copy()

goolamn_metadata['Comment[ENA_RUN]'].to_csv("../pipeline/fetchngs/E-MTAB-3321.txt", index=None, header=None)
goolamn_metadata.set_index('Comment[ENA_RUN]').to_csv("../pipeline/fetchngs/E-MTAB-3321_metadata.csv")

1.4.2 Load dataset

goolam = load_experiment(
    "../data/external/aligned/mouse/Goolam_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="E-MTAB-3321", left_on='run_accession', right_on='Comment[ENA_RUN]'
)

goolam.obs['batch'] = [f"GOOLAM_{batch}" for batch in goolam.obs["Characteristics[batch]"].values]
goolam.obs['experiment'] = "Goolamn et al., 2016"
goolam.obs['technology'] = "SMART-seq2"

goolam.obs = goolam.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

goolam = normalize_smartseq(goolam, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

goolam

AnnData object with n_obs × n_vars = 118 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.5 Boroviak et al., 2015 [E-MTAB-2958] and [E-MTAB-2959]

nf-core_tower.sh \
    Boroviak_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/E-MTAB-29589.txt

nf-core_tower.sh Boroviak_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Boroviak_et_al/results/samplesheet/samplesheet.csv

1.5.1 Preparation

boroviak_metadata = pd.concat([
    pd.read_table("https://ftp.ebi.ac.uk/biostudies/nfs/E-MTAB-/958/E-MTAB-2958/Files/E-MTAB-2958.sdrf.txt"),
    pd.read_table("https://ftp.ebi.ac.uk/biostudies/nfs/E-MTAB-/959/E-MTAB-2959/Files/E-MTAB-2959.sdrf.txt")
])

boroviak_metadata['ct'] = boroviak_metadata['Extract Name'].values

legend = {
    'E2.5_MOR': '8C',
    'E3.5_ICM': 'E3.5-ICM',
    'E4.5_PrE': 'E4.5-PrE',
    'E4.5_EPI': 'E4.5-EPI',
}

for old_name, new_name in legend.items():
    boroviak_metadata.loc[boroviak_metadata.ct.str.startswith(old_name), 'ct'] = new_name

boroviak_metadata = boroviak_metadata.query('ct in ["8C", "E3.5-ICM", "E4.5-EPI", "E4.5-PrE"]').copy()
boroviak_metadata.ct = boroviak_metadata.ct.astype('category')

boroviak_metadata['Comment[ENA_RUN]'].to_csv("../pipeline/fetchngs/E-MTAB-29589.txt", index=None, header=None)
boroviak_metadata.set_index('Comment[ENA_RUN]').to_csv("../pipeline/fetchngs/E-MTAB-29589_metadata.csv")

1.5.2 Load dataset

boroviak = load_experiment(
    "../data/external/aligned/mouse/Boroviak_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="E-MTAB-29589", left_on='run_accession', right_on='Comment[ENA_RUN]')

boroviak.obs['batch'] = "BOROVIAK_1"
boroviak.obs['experiment'] = "Boroviak et al., 2015"
boroviak.obs['technology'] = "SMART-seq2"

boroviak.obs = boroviak.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

boroviak = normalize_smartseq(boroviak, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

boroviak

AnnData object with n_obs × n_vars = 12 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.6 Chen et al., 2016 GSE74155

nf-core_tower.sh \
    Chen_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --force_sratools_download \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE74155.txt

nf-core_tower.sh Chen_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Chen_et_al/results/samplesheet/samplesheet.csv

1.6.1 Preparation

chen_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE74nnn/GSE74155/matrix/GSE74155_series_matrix.txt.gz",
                              skiprows=30, index_col=0).T

chen_metadata['ct'] = chen_metadata['!Sample_characteristics_ch1'].iloc[:, 1].values
chen_metadata['SRX'] = chen_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')
chen_metadata = chen_metadata.reset_index()

legend = {
    'cell type: E3.5 ICM': 'E3.5-ICM',
    'cell type: E4.5 Epiblast': 'E4.5-EPI'
}

for old_name, new_name in legend.items():
    chen_metadata.loc[chen_metadata.ct.str.startswith(old_name), 'ct'] = new_name

chen_metadata = chen_metadata.query('ct in ["E3.5-ICM", "E4.5-EPI"]').copy()
chen_metadata.ct = chen_metadata.ct.astype('category')

chen_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE74155.txt", index=None, header=None)
chen_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE74155_metadata.csv")

1.6.2 Load dataset

chen = load_experiment(
    "../data/external/aligned/mouse/Chen_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE74155", left_on='sample_alias', right_on='ID_REF')

chen.obs['batch'] = "CHEN_1"
chen.obs['experiment'] = "Chen et al., 2016"
chen.obs['technology'] = "SMART-seq2"

chen.obs = chen.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

chen = normalize_smartseq(chen, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

chen

AnnData object with n_obs × n_vars = 46 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.7 Nowotschin et al., 2019 [GSE123046]

nf-core_tower.sh \
    Nowotschin_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --force_sratools_download \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE123046.txt

wget https://raw.githubusercontent.com/10XGenomics/cellranger/master/lib/python/cellranger/barcodes/737K-august-2016.txt ../pipeline/737K-august-2016.txt && gzip ../pipeline/737K-august-2016.txt

nf-core_tower.sh Nowotschin_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/10X.config \
    --input /scratch/Brickman/pipelines/Nowotschin_et_al/results/samplesheet/samplesheet.csv

1.7.1 Preparation

nowotschin_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE123nnn/GSE123046/matrix/GSE123046_series_matrix.txt.gz",
                                    skiprows=41, index_col=0).T

nowotschin_metadata['ct'] = nowotschin_metadata.index
nowotschin_metadata['SRX'] = nowotschin_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')
nowotschin_metadata = nowotschin_metadata.reset_index()

nowotschin_metadata = nowotschin_metadata[nowotschin_metadata.ct.str.startswith(("E3", "E4"))].copy()

nowotschin_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE123046.txt", index=None, header=None)
nowotschin_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE123046_metadata.csv")

1.7.2 Load dataset

def nowotschin_get_metadata():
    metadata = pd.read_csv("https://s3.amazonaws.com/dp-lab-data-public/mouse_endoderm/sc_endoderm_all_cells_metadata.csv", index_col=0)
    correction = pd.read_csv("../data/external/aligned/mouse/Nowotschin_et_al/e35_cell_types.csv", index_col=0)

    metadata.loc[correction.index, 'CellType'] = correction.CellType

    return metadata

nowotschin = sc.read("../data/external/aligned/mouse/Nowotschin_et_al/results/star/mtx_conversions/combined_matrix.h5ad")

legend = {
    'E4.5_Rep1': 'Lib1-1_E4.5',
    'E4.5_Rep2': 'Lib1-2_E4.5',
    'E3.5_Rep1': 'Lib1-3_E3.5',
    'E3.5_Rep2': 'Lib1-4_E3.5',
}

nowotschin.obs['Library'] = 'UNKNOWN'
for old_name, new_name in legend.items():
    nowotschin.obs.loc[nowotschin.obs['sample_title'] == old_name, 'Library'] = new_name

nowotschin.obs['cell_id'] = nowotschin.obs['Library'] + "_" + nowotschin.obs_names.str.split('_', expand=True).droplevel(1)
nowotschin.obs['timepoint'] = nowotschin.obs.sample_title.str.split('_', expand=True).values[:, 0]
nowotschin.obs = nowotschin.obs.set_index('cell_id')

nowotschin_metadata = nowotschin_get_metadata()

nowotschin = nowotschin[nowotschin.obs_names.intersection(nowotschin_metadata.index)].copy()
nowotschin_metadata = nowotschin_metadata.loc[nowotschin.obs_names].copy()

nowotschin.obs['batch'] = "NOWO_1"
nowotschin.obs.loc[nowotschin.obs['timepoint'].str.startswith('E4.5'), 'batch'] = "NOWO_2"
# nowotschin.obs['batch'] = nowotschin.obs['Library'].replace({
#     'Lib1-1_E4.5': 'NOWO_1', 
#     'Lib1-2_E4.5': 'NOWO_2',
#     'Lib1-3_E3.5': 'NOWO_3',
#     'Lib1-4_E3.5': 'NOWO_4', 
# })
nowotschin.obs['experiment'] = "Nowotschin et al., 2019"
nowotschin.obs['technology'] = "10X 3' v2"
nowotschin.obs['ct'] = nowotschin_metadata[['Timepoint', 'CellType']].agg('-'.join, axis=1)
nowotschin.obs['ct_orig'] = nowotschin_metadata['CellType']

nowotschin.obs = nowotschin.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

1.8 Fan et al., 2015 GSE53386

nf-core_tower.sh \
    Fan_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --force_sratools_download \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE53386.txt

nf-core_tower.sh Fan_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Fan_et_al/results/samplesheet/samplesheet.csv

1.8.1 Preparation

fan_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE53nnn/GSE53386/matrix/GSE53386-GPL13112_series_matrix.txt.gz", 
                      skiprows=33, index_col=0).T
fan_metadata['SRX'] = fan_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')

fan_metadata['ct'] = fan_metadata['!Sample_characteristics_ch1'].iloc[:, 0].values
fan_metadata = fan_metadata.reset_index()

legend = {
    'cell type: zygote': 'Zygote',
    'cell type: 2-cell embryo': '2C',
    'cell type: 4-cell embryo': '4C',
    'cell type: 8-cell embryo': '8C', 
    'cell type: morula': '16C',
}

for old_name, new_name in legend.items():
    fan_metadata.loc[fan_metadata.ct == old_name, 'ct'] = new_name
    
fan_metadata = fan_metadata[fan_metadata.ct.isin(legend.values())].copy()

fan_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE53386.txt", index=None, header=None)
fan_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE53386_metadata.csv")

1.8.2 Load dataset

fan = load_experiment(
    "../data/external/aligned/mouse/Fan_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE53386"
)

fan.obs['batch'] = "FAN_1"
fan.obs['experiment'] = "Fan et al., 2015"
fan.obs['technology'] = "SUPeR-seq"

fan.obs = fan.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

fan = normalize_smartseq(fan, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

fan

AnnData object with n_obs × n_vars = 27 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.9 Borensztein et al., 2017 GSE80810

nf-core_tower.sh \
    Borensztein_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE80810.txt

nf-core_tower.sh Borensztein_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Borensztein_et_al/results/samplesheet/samplesheet.csv

1.9.1 Preparation

borenszrtein_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE80nnn/GSE80810/matrix/GSE80810_series_matrix.txt.gz",
                                      skiprows=31, index_col=0).T
borenszrtein_metadata['SRX'] = borenszrtein_metadata[['!Sample_relation']].iloc[:, :2].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')

borenszrtein_metadata['ct'] = borenszrtein_metadata['!Sample_characteristics_ch1'].iloc[:, 1].values
borenszrtein_metadata = borenszrtein_metadata.reset_index()

legend = {
    # 'developmental stage: Oocyte': 'Oocyte',
    'developmental stage: 2-cell': '2C',
    'developmental stage: 4-cell': '4C',
    'developmental stage: 8-cell': '8C',
    'developmental stage: 16-cell': '16C',
    # 'developmental stage: 32-cell': 'UNKNOWN',
    # 'developmental stage: 64-cell': 'UNKNOWN'
}

for old_name, new_name in legend.items():
    borenszrtein_metadata.loc[borenszrtein_metadata.ct == old_name, 'ct'] = new_name
    
borenszrtein_metadata = borenszrtein_metadata[borenszrtein_metadata.ct.isin(legend.values())].copy()
borenszrtein_metadata = borenszrtein_metadata[borenszrtein_metadata['!Sample_characteristics_ch1'].iloc[:, 2].str.contains('wt/wt')].copy()

borenszrtein_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE80810.txt", index=None, header=None)
borenszrtein_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE80810_metadata.csv")

1.9.2 Load dataset

borensztein = load_experiment(
    "../data/external/aligned/mouse/Borensztein_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE80810", left_on='sample_alias', right_on='ID_REF'
)

borensztein.obs['batch'] = "BORENSZTEIN_1"
borensztein.obs['experiment'] = "Borensztein et al., 2017"
borensztein.obs['technology'] = "qRT-PCR"

borensztein.obs = borensztein.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

borensztein = normalize_smartseq(borensztein, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

borensztein

AnnData object with n_obs × n_vars = 60 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.10 Stirparo et al., 2020 GSE159030

nf-core_tower.sh \
    Stirparo_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --force_sratools_download \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE159030.txt

nf-core_tower.sh Stirparo_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Stirparo_et_al/results/samplesheet/samplesheet.csv

1.10.1 Preparation

stirparo_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE159nnn/GSE159030/matrix/GSE159030_series_matrix.txt.gz",
                                  skiprows=27, index_col=0).T
stirparo_metadata['SRX'] = stirparo_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')

stirparo_metadata['ct'] = stirparo_metadata['!Sample_source_name_ch1']
stirparo_metadata = stirparo_metadata.reset_index()

legend = {
    'Primitive Endoderm': 'PrE',
    'Epiblast': 'EPI',
}

for old_name, new_name in legend.items():
    stirparo_metadata.loc[stirparo_metadata.ct == old_name, 'ct'] = new_name

stirparo_metadata['ct'] = stirparo_metadata['!Sample_characteristics_ch1'].iloc[:, 1].str.replace('developmental stage: ', '').astype(str)\
                        + '-' + stirparo_metadata.ct
stirparo_metadata = stirparo_metadata[stirparo_metadata['!Sample_characteristics_ch1'].iloc[:, 0] == 'genotype: Wild-Type'].copy()

stirparo_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE159030.txt", index=None, header=None)
stirparo_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE159030_metadata.csv")

1.10.2 Load dataset

stirparo = load_experiment(
    "../data/external/aligned/mouse/Stirparo_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE159030", left_on='sample_alias', right_on='ID_REF'
)

stirparo.obs['batch'] = "STIRPARO_1"
stirparo.obs['experiment'] = "Stirparo et al., 2020"
stirparo.obs['technology'] = "SMART-seq2"

stirparo.obs = stirparo.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

stirparo = normalize_smartseq(stirparo, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

stirparo

AnnData object with n_obs × n_vars = 64 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.11 Xue et al., 2013 GSE44183

nf-core_tower.sh \
    Xue_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE44183.txt

nf-core_tower.sh Xue_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Xue_et_al/results/samplesheet/samplesheet.csv

1.11.1 Preparation

xue_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE44nnn/GSE44183/matrix/GSE44183-GPL13112_series_matrix.txt.gz",
                               skiprows=35, index_col=0).T
xue_metadata['SRX'] = xue_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')

xue_metadata['ct'] = xue_metadata['!Sample_source_name_ch1']
xue_metadata = xue_metadata.reset_index()

legend = {
    # 'oocyte', : '',
    # 'pronucleus': '',
    '2-cell blastomere': '2C', 
    '4-cell blastomere': '4C',
    '8-cell blastomere': '8C', 
    'morula': '16C'
}

for old_name, new_name in legend.items():
    xue_metadata.loc[xue_metadata.ct == old_name, 'ct'] = new_name

xue_metadata = xue_metadata[xue_metadata.ct.isin(legend.values())].copy()

xue_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE44183.txt", index=None, header=None)
xue_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE44183_metadata.csv")

1.11.2 Load dataset

xue = load_experiment(
    "../data/external/aligned/mouse/Xue_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE44183"
)

xue.obs['batch'] = "XUE_1"
xue.obs['experiment'] = "Xue et al., 2013"
xue.obs['technology'] = "qRT-PCR"

xue.obs = xue.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

xue = normalize_smartseq(xue, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

xue

AnnData object with n_obs × n_vars = 12 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

1.12 Yanagida et al., 2022 GSE148462

nf-core_tower.sh \
    Yanagida_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --force_sratools_download \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE148462.txt

nf-core_tower.sh Yanagida_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Yanagida_et_al/results/samplesheet/samplesheet.csv

1.12.1 Preparation

yanagida_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE148nnn/GSE148462/matrix/GSE148462_series_matrix.txt.gz",
                                  skiprows=29, index_col=0).T
yanagida_metadata['SRX'] = yanagida_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')

yanagida_metadata['ct'] = 'E3.75-ICM'

yanagida_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE148462.txt", index=None, header=None)
yanagida_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE148462_metadata.csv")

1.12.2 Load dataset

yanagida = load_experiment(
    "../data/external/aligned/mouse/Yanagida_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE148462", left_on='sample_alias', right_on='ID_REF'
)

yanagida.obs['batch'] = "YANAGIDA_1"
yanagida.obs['experiment'] = "Yanagida et al., 2022"
yanagida.obs['technology'] = "SMART-seq2"

yanagida.obs = yanagida.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

yanagida = normalize_smartseq(yanagida, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

yanagida

AnnData object with n_obs × n_vars = 48 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

2 Mohammed et al., 2017 GSE100597

nf-core_tower.sh \
    Mohammed_et_al \
    nextflow run nf-core/fetchngs \
    -r 1.10.0 \
    --input /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/fetchngs/GSE100597.txt

nf-core_tower.sh Mohammed_et_al nextflow run brickmanlab/scrnaseq \
    -r feature/smartseq \
    -c /projects/dan1/data/Brickman/projects/proks-salehin-et-al-2023/pipeline/smartseq.config \
    --input /scratch/Brickman/pipelines/Mohammed_et_al/results/samplesheet/samplesheet.csv

2.1 Preparation

mohammed_metadata = pd.read_table("https://ftp.ncbi.nlm.nih.gov/geo/series/GSE100nnn/GSE100597/matrix/GSE100597_series_matrix.txt.gz",
                                  skiprows=27, index_col=0).T
mohammed_metadata['SRX'] = mohammed_metadata[['!Sample_relation']].agg(' '.join, axis=1).str.extract(r'(SRX[0-9]{6})')

mohammed_metadata = mohammed_metadata[mohammed_metadata['!Sample_source_name_ch1'].isin(['E3.5', 'E4.5'])]
mohammed_metadata = mohammed_metadata[mohammed_metadata.index.str.endswith('single')].copy()

mohammed_metadata['!Sample_geo_accession'].to_csv("../pipeline/fetchngs/GSE100597.txt", index=None, header=None)
mohammed_metadata.set_index('SRX').to_csv("../pipeline/fetchngs/GSE100597_metadata.csv")

2.2 Load data

mohammed = load_experiment(
    "../data/external/aligned/mouse/Mohammed_et_al/results/star/mtx_conversions/combined_matrix.h5ad",
    GEO="GSE100597", left_on='sample_alias', right_on='ID_REF'
)

mohammed_metadata = pd.read_table("../data/external/aligned/mouse/Mohammed_et_al/Cell_clusters.txt", index_col=0)
mohammed_metadata.Lineage = mohammed_metadata.Lineage.str.replace('epiblast', 'EPI').str.replace('PE', 'PrE')
mohammed.obs['ct'] = mohammed_metadata.loc[mohammed.obs.ct_orig, ['Stage', 'Lineage']].agg('-'.join, axis=1).values
mohammed = mohammed[mohammed.obs.ct != 'E4.5-interm'].copy()

mohammed.obs['batch'] = "MOHAMMED_1"
mohammed.obs['experiment'] = "Mohammed et al., 2017"
mohammed.obs['technology'] = "SMART-seq2"

mohammed.obs = mohammed.obs[['batch', 'experiment', 'technology', 'ct', 'ct_orig']]

mohammed = normalize_smartseq(mohammed, genes_length)

SMART-SEQ: Normalization
SMART-SEQ: Common genes 55364

mohammed

AnnData object with n_obs × n_vars = 153 × 55364
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig'
    var: 'gene_symbol'

3 Integration

import urllib.request, json

cc_url = "https://github.com/brickmanlab/project-template/raw/master/%7B%7B%20cookiecutter.project_name%20%7D%7D/data/external/mouse_cell_cycle_genes.json"
with urllib.request.urlopen(cc_url) as url:
    cc_dict = json.load(url)
    cc = sum(list(cc_dict.values()), [])

adata = anndata.concat([deng, biase, posfai, goolam, boroviak, chen, nowotschin, fan, borensztein, stirparo, xue, yanagida, mohammed])
adata.obs['stage'] = adata.obs.ct.str.replace('E3.25-|E3.5-|E3.75-|E4.5-', '', regex=True)
adata.obs['timepoint'] = adata.obs.ct.str.split('-', expand=True).iloc[:, 0]
adata.obs.ct = pd.Categorical(adata.obs.ct, 
                              categories=['Zygote', '2C', '4C', '8C', '16C', 
                                          'E3.25-ICM', 'E3.25-TE', 
                                          'E3.5-ICM', 'E3.5-TE', 'E3.5-EPI', 'E3.5-PrE', 
                                          'E3.75-ICM', 
                                          'E4.5-TE', 'E4.5-EPI', 'E4.5-PrE'],
                              ordered=True)
adata.obs_names_make_unique()

adata.var['gene_ids'] = adata.var_names
adata.var['gene_symbol'] = deng[:, adata.var_names].var['gene_symbol'].values
adata.var.index = adata.var.gene_symbol.str.lower().values
adata.var_names_make_unique()

# remove ribosomal genes
adata = adata[:, ~np.logical_or(adata.var_names.str.startswith(('rps','rpl')), adata.var_names == "ct010467.1")].copy()

# remove cc genes
adata = adata[:, ~adata.var_names.isin(cc)].copy()

adata

AnnData object with n_obs × n_vars = 2022 × 54940
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig', 'stage', 'timepoint'
    var: 'gene_ids', 'gene_symbol'

get_summary(adata)

Number of MITO genes: 0
Number of ERCC genes: 0
Number of RIBO genes: 0

3.1 Number of cells per experiment

/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/seaborn/categorical.py:82: FutureWarning: iteritems is deprecated and will be removed in a future version. Use .items instead.
  plot_data = [np.asarray(s, float) for k, s in iter_data]

3.2 Number of cells cell type

3.3 QC plots

4 Final adjustments

adata

AnnData object with n_obs × n_vars = 2022 × 54940
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig', 'stage', 'timepoint', 'n_genes_by_counts', 'total_counts', 'total_counts_mt', 'pct_counts_mt'
    var: 'gene_ids', 'gene_symbol', 'mt', 'n_cells_by_counts', 'mean_counts', 'pct_dropout_by_counts', 'total_counts'

sc.pl.violin(adata, ['n_genes_by_counts', 'total_counts', 'pct_counts_mt'], jitter=0.4, multi_panel=True)

sc.pp.filter_genes(adata, min_cells=10)
sc.pp.filter_cells(adata, max_genes=20_000)
sc.pp.filter_cells(adata, max_counts=26_000_000)

# adata = adata[adata.obs.pct_counts_mt < 30]

sc.pl.violin(adata, ['n_genes_by_counts', 'total_counts', 'pct_counts_mt'], jitter=0.4, multi_panel=True)

fig, ax = plt.subplots(figsize=(20, 5), sharey=True)
sns.scatterplot(
    x="total_counts", y="n_genes_by_counts", ax=ax, data=adata.obs, hue="ct"
)
# plt.axvline(x=26_000_000, c='r', linestyle='--')

<Axes: xlabel='total_counts', ylabel='n_genes_by_counts'>

adata.layers["counts"] = adata.X.copy()

sc.pp.normalize_total(adata)
sc.pp.log1p(adata)
adata.raw = adata

adata

AnnData object with n_obs × n_vars = 2004 × 34346
    obs: 'batch', 'experiment', 'technology', 'ct', 'ct_orig', 'stage', 'timepoint', 'n_genes_by_counts', 'total_counts', 'total_counts_mt', 'pct_counts_mt', 'n_genes', 'n_counts'
    var: 'gene_ids', 'gene_symbol', 'mt', 'n_cells_by_counts', 'mean_counts', 'pct_dropout_by_counts', 'total_counts', 'n_cells'
    uns: 'log1p'
    layers: 'counts'

adata.write("../data/processed/01_mouse_reprocessed.h5ad")

sc.tl.pca(adata, svd_solver='arpack')
sc.pl.pca(adata, color=['ct', 'batch', 'technology', 'experiment'], frameon=False, wspace=0.4, ncols=2)

/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/scanpy/plotting/_tools/scatterplots.py:392: UserWarning: No data for colormapping provided via 'c'. Parameters 'cmap' will be ignored
  cax = scatter(
/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/scanpy/plotting/_tools/scatterplots.py:392: UserWarning: No data for colormapping provided via 'c'. Parameters 'cmap' will be ignored
  cax = scatter(
/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/scanpy/plotting/_tools/scatterplots.py:392: UserWarning: No data for colormapping provided via 'c'. Parameters 'cmap' will be ignored
  cax = scatter(
/projects/dan1/data/Brickman/conda/envs/scvi-1.0.0/lib/python3.10/site-packages/scanpy/plotting/_tools/scatterplots.py:392: UserWarning: No data for colormapping provided via 'c'. Parameters 'cmap' will be ignored
  cax = scatter(