ScEasyMode: Wrappers for automating single cell workflows in Python
Features
- Multiseq correction in python using a within-barcode zscore correction
- Plotting for stacked barplots in your dataset
- Mouse cell filtering/separation from mixed dataset
- Scanpy wrapper that simplifies the workflow
Installation
Install using Pip
pip3 install git+https://github.com/johnnyUCSF/scEasyMode#egg='scEasyMode'Install using a Conda Environment
- You may also use Conda to start an environment with ScEasyMode installed inside it. You can install conda from here.
- Firstly, clone the repository and create the environment as shown below. Then, activate the environment.
git clone https://github.com/johnnyUCSF/scEasyMode
cd scEasyMode
conda env create -f environment.yml
conda activate sceasymode_env- Now start your Jupyter Notebook or Python shell inside the conda environment
Usage
Load the modules
from scEasyMode import mousefilter
from scEasyMode import clusterplot
from scEasyMode import pymulti
from scEasyMode import sceasyDemultiplex your samples
import pandas as pd
from scEasyMode import pymulti
# Define parameters
len_10x=16 # Number of bases in cell barcodes
len_umi=12 # Length of UMI
len_multi=15, # Number of bases in the HTO barcodes / HashTag O
fastq_r1 = 'path/to/file'
fastq_r2 = 'path/to/file'
sample_name = 'test_demultiplexing'
cell_BC_file = 'path/to/cell_barcodes' # Counts Matrix after alignment and pre-processing
cell_bcs = pd.read_csv(cell_BC_file, sep='\t', header=None)[0].tolist()
multi_BC_file = 'path/to/barcodes' # Barcodes TSV file from 10x or Illumina
bcsmulti = pd.read_csv(multi_BC_file,sep=',',index_col=1,header=None)
bcsmulti.columns = ['multi']
bcsmulti = bcsmulti['multi'].tolist()
pymulti.pymulti(fastq_r1, fastq_r2, bcsmulti=bcsmulti, bcs10x=cell_bcs,
len_10x=len_10x, len_multi=len_multi, len_umi=len_umi, split=True,
hamming=True, median_only=True, sampname= sample_name, filter_unmapped_reads=True)
# This function will output multiple graphs
# It will also store a matrix of the assigned barcodes in the 'pymulti' directory inside the working directory.
# Note that some reads are unmapped. If you want to retain them, you can do so by specifying filter_unmapped_reads=False.
