cellitac 1.0.3

Cell type Identification using Transcription factor Analysis and Chromatin accessibility

cellitac

Cell type Identification using Transcription factor Analysis and Chromatin accessibility

A pipeline for processing Single-Cell ATAC + RNA Multiome data and classifying cell types using Machine Learning.

---

What It Does

Stage	Steps	Tools
Preprocessing	RNA QC → normalization → cell-type annotation	Seurat + SingleR (R via rpy2)
Preprocessing	ATAC QC → TF-IDF → LSI	Signac (R via rpy2)
Preprocessing	RNA + ATAC integration → ML-ready CSVs	Pure Python
ML	Imbalance analysis → SMOTE → feature selection	scikit-learn, imbalanced-learn
ML	RF + XGBoost + SVM training & evaluation	scikit-learn, xgboost
ML	19 plots + JSON report + XLSX	matplotlib, seaborn, networkx

---

Requirements

Before installing cellitac, you need:

• Linux or macOS (Ubuntu 20.04+ recommended)
• Python 3.9, 3.10, or 3.11 (not 3.12 or higher)
• Conda / Miniconda (download here)
• ~5 GB free disk space

---

Installation

Step 1 — Create a Conda environment

conda create -n cellitac python=3.11 -y
conda activate cellitac

Step 2 — Install R and core R libraries via conda

conda install -c conda-forge r-base=4.3.1 -y

conda install -c conda-forge -c bioconda \
  r-matrix r-hdf5r rpy2 \
  bioconductor-summarizedexperiment \
  bioconductor-singlecellexperiment \
  bioconductor-genomicranges \
  bioconductor-delayedarray \
  bioconductor-biocsingular \
  bioconductor-biocneighbors \
  bioconductor-genomicalignments \
  bioconductor-genomicfeatures \
  bioconductor-rtracklayer -y

Step 3 — Install remaining R packages (takes 10–30 min)

Rscript -e "install.packages('BiocManager', repos='https://cran.r-project.org')"

Rscript -e "BiocManager::install(c(
  'Seurat', 'Signac', 'SingleR', 'celldex',
  'EnsDb.Hsapiens.v75', 'biovizBase', 'data.table'
), ask=FALSE)"

Step 4 — Install cellitac

pip install cellitac

Step 5 — Verify installation

cellitac --help

If you see the help message, you are ready to go ✅

---

Quick Start

Download test data (PBMC 3k cells, ~560 MB)

mkdir -p ~/data && cd ~/data

wget https://cf.10xgenomics.com/samples/cell-arc/2.0.0/pbmc_granulocyte_sorted_3k/pbmc_granulocyte_sorted_3k_filtered_feature_bc_matrix.h5
wget https://cf.10xgenomics.com/samples/cell-arc/2.0.0/pbmc_granulocyte_sorted_3k/pbmc_granulocyte_sorted_3k_atac_fragments.tsv.gz
wget https://cf.10xgenomics.com/samples/cell-arc/2.0.0/pbmc_granulocyte_sorted_3k/pbmc_granulocyte_sorted_3k_atac_fragments.tsv.gz.tbi
wget https://cf.10xgenomics.com/samples/cell-arc/2.0.0/pbmc_granulocyte_sorted_3k/pbmc_granulocyte_sorted_3k_atac_peaks.bed
wget https://cf.10xgenomics.com/samples/cell-arc/2.0.0/pbmc_granulocyte_sorted_3k/pbmc_granulocyte_sorted_3k_per_barcode_metrics.csv

Run the pipeline

conda activate cellitac
cellitac --input ~/data --output ~/results

---

Full Dataset (PBMC 10k)

mkdir -p ~/data && cd ~/data

wget https://cf.10xgenomics.com/samples/cell-arc/1.0.0/pbmc_unsorted_10k/pbmc_unsorted_10k_filtered_feature_bc_matrix.h5
wget https://cf.10xgenomics.com/samples/cell-arc/1.0.0/pbmc_unsorted_10k/pbmc_unsorted_10k_per_barcode_metrics.csv
wget https://cf.10xgenomics.com/samples/cell-arc/1.0.0/pbmc_unsorted_10k/pbmc_unsorted_10k_atac_fragments.tsv.gz
wget https://cf.10xgenomics.com/samples/cell-arc/1.0.0/pbmc_unsorted_10k/pbmc_unsorted_10k_atac_fragments.tsv.gz.tbi
wget https://cf.10xgenomics.com/samples/cell-arc/1.0.0/pbmc_unsorted_10k/pbmc_unsorted_10k_atac_peaks.bed

Note: cellitac auto-detects file names — your files do not need to follow the 10x naming convention.

---

Usage

Command Line

# Full pipeline (preprocessing + ML)
cellitac --input ~/data --output my_results

# Preprocessing only
cellitac-preprocess --input ~/data --output my_results

# ML only (if preprocessing already done)
cellitac-model --data my_results/python_ready_data --output my_results/ml

Python API

from cellitac import run_full_pipeline, run_preprocessing, run_model

# Full pipeline
run_full_pipeline(input_dir="~/data", output_dir="my_results")

# Preprocessing only
run_preprocessing(input_dir="~/data", output_dir_python="python_ready_data")

# ML only
run_model(data_dir="python_ready_data", output_dir="ml_results")

# Use the ML class directly
from cellitac.mainModel import scATACMLPipeline
pipeline = scATACMLPipeline(data_dir="python_ready_data", output_dir="ml_results")
pipeline.run_complete_pipeline()

---

Input Files

File	Extension	Required
Feature-barcode matrix	.h5	✅ Yes
ATAC fragments	.tsv.gz	✅ Yes
Fragments index	.tsv.gz.tbi	✅ Yes
Peaks BED file	.bed	✅ Yes
Per-barcode QC metrics	.csv	⭕ Optional

---

Output Files

File	Description
ml_pipeline_report.json	Full JSON report
model_performance_summary.csv	Accuracy / F1 / AUC per model
detailed_model_results.xlsx	Per-class metrics, CV results
model_performance_comparison.png	Bar chart comparison
confusion_matrices.png	Confusion matrices
class_distribution_analysis.png	Cell type distribution
class_balancing_comparison.png	Before/after SMOTE
feature_importance.png	RF + XGBoost top 20 features
simple_feature_heatmap.png	Feature importance heatmap
overfitting_analysis.png	CV train vs validation
learning_curves.png	Learning curves per model
performance_radar.png	Radar chart
feature_distributions.png	Violin plots
class_separation_pca.png	PCA scatter
basic_tf_network.png	Feature–cell-type network

---

Package Structure

cellitac/
├── src/cellitac/
│   ├── __init__.py          # Public API
│   ├── config.py            # Parameters (paths, QC thresholds, ML hyperparams)
│   ├── pipeline.py          # run_preprocessing, run_model, run_full_pipeline
│   ├── preprocessing.py     # R preprocessing via rpy2
│   ├── mainModel.py         # scATACMLPipeline class (19-step ML pipeline)
│   ├── cli.py               # cellitac / cellitac-preprocess / cellitac-model
│   └── rscripts/
│       ├── team1_rna.R      # Seurat + SingleR
│       └── team2_atac.R     # Signac
├── tests/
│   └── test_model.py
├── pyproject.toml
└── README.md

---

Troubleshooting

Problem	Solution
conda activate cellitac not working	Run conda init then restart terminal
R packages fail to install	Make sure you installed from conda first (Step 2) before BiocManager (Step 3)
hdf5r error	Run conda install -c conda-forge hdf5 r-hdf5r -y
peak_region_fragments not found	Normal for some datasets — pipeline continues automatically
slot deprecated error	Make sure you have the latest cellitac version: pip install --upgrade cellitac

---

Tests

pip install cellitac[dev]
pytest tests/ -v

---

Contributors

📧 1. Rana H. Abu-Zeid — [e-mail] 📧 2. Syrus Semawule — [e-mail] 📧 3. Emmanuel Aroma — [e-mail] 📧 4. Toheeb Jumah — [e-mail] 📧 5. Derek Reiman, Ph.D. — [e-mail] 📧 6. Olaitan I. Awe, Ph.D. — [e-mail]

---

License

MIT